Practical Exploration of Data-Driven Strategy Optimisation of E-commerce Platforms to Promote Regional Common Prosperity
Publié en ligne: 05 févr. 2025
Reçu: 19 sept. 2024
Accepté: 08 janv. 2025
DOI: https://doi.org/10.2478/amns-2025-0061
Mots clés
© 2025 Yijing Huang, published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.
In terms of income, “affluent” means that people’s income level has increased, and “sharing” has manifested itself in the narrowing of the income gap between regions and urban and rural areas [1-2]. E-commerce helps to consolidate the material foundation of the common wealth, which not only promotes the high-level development and utilization of regional data elements, improves the factor mismatch, optimizes the allocation of factor resources, and promotes efficiency changes, but also promotes the extension of regional industrial chain, gives play to the scale advantage effect of industrial agglomeration, and promotes the economy to achieve the qualitative and quantitative effective improvement and reasonable growth [3-5]. The impact of e-commerce on “sharing” has both positive and negative effects. From the perspective of positive effects, e-commerce gives full play to the digital dividend between regions, improves the unbalanced development status quo, ensures the bottom line standard of common prosperity, promotes the concentration of high-level talents, and exerts the universal effect of human capital [6-8]. However, in terms of negative effects, the regional enhancement dividend brought by e-commerce has widened the urban-rural gap, widened the regional development gap, caused uneven income distribution among different groups, and although it has raised the level of residents’ income in general, it has a greater effect on the enhancement of the families with a higher level of education and rich social capital [9-12]. In this context, scientific investigation of the impact effect of e-commerce on the common wealth and its mechanism of action can not only enrich the relevant research on the relationship between the digital economy and the common wealth but also provide a basis for the effective use of e-commerce means to empower the common wealth in the new journey [13-15].
The digital economy, as a new type of economic form, has an increasingly prominent impact on economic and social development and is becoming an important engine for promoting more visible and substantial progress in common prosperity [16]. Liu, W. et al. analyzed in detail the integration and development path of the digital economy and rural economy. On the one hand, the use of digital finance to develop the rural economy, reduce the gap between urban and rural economies, and promote common prosperity. On the other hand, the excellent rural traditional culture in the digital culture industry opens up a new mode of traditional culture development [17]. Jiechang, X. et al. clarified that the common wealth inevitably relies on the digital economy, which not only realizes the general growth of the macro-economy, but also realizes the balanced growth of the regional economy, and at the same time, accelerates the equalization of public services, but also pays attention to the structural unemployment and other problems brought about by the rise of the cause of the common wealth and promotes the sustainable and stable development of the cause of the common wealth [18]. Chen, L. et al. explored the mechanism of the role of the digital economy on the common wealth, pointing out that the promotion of the digital economy on the common wealth has dynamic and non-linear characteristics and has a significant spatial spillover effect, while resource allocation plays a mediating role in this promotion process [19].
E-commerce, as a form of digital economy, plays an equally important role in regional common prosperity. Wei, X. et al. studied the impact of China’s comprehensive demonstration policy of e-commerce in agriculture on urban and rural common prosperity and found that e-commerce promotes regional common prosperity while there is a discrepancy between the eastern region, the central region and the western region [20]. Jie, X. U. et al. empirically investigated the promotion effect of e-commerce development on the level of common wealth in counties and found that policy effect, human capital level, and county governance capacity all have significant positive impacts on the common wealth in counties in addition to the fact that there are regional developmental differences in common wealth [21]. Cheng, J. et al. found that rural e-commerce promoted residents’ material affluence, in which entrepreneurship, social capital and financial literacy played a positive moderating role, as well as spiritual affluence and ecological affluence, based on the analysis of a large number of sample data [22].
In this study, the “Comprehensive Demonstration Policy of E-commerce in Rural Areas”, which was implemented in 2014, is regarded as a “quasi-natural experiment” of data-driven strategy optimization of e-commerce platforms based on sample data from 204 prefecture-level cities in China from 2014 to 2023. The urban-rural income gap was selected as the explanation variable. The level of economic development of the e-commerce platform was taken as the main explanatory variable. The dual machine learning model was used to explore the effect of data-driven strategy optimization of e-commerce platforms on regional common prosperity. After the parallel trend test and the placebo test, the heterogeneity analysis is carried out from the aspects of regional characteristics and the external environment so as to provide constructive suggestions for the rapid and steady development of e-commerce platforms and the promotion of regional common prosperity.
The common wealth of the region is important in eliminating the disparity between the rich and the poor in rural and urban areas, and e-commerce is a typical practice of the deep integration of digital technology and the real economy, as well as a breakthrough in the development of the rural digital economy. In recent years, the application of e-commerce in China’s rural areas has gradually expanded and deepened, with far-reaching impacts on industrial development, production, and business models in rural areas. The number of Taobao villages and the per capita disposable income of rural residents are shown in Figure 1. Taking Taobao villages as an example, it has increased from 100 in 2014 to 9,136 in 2023, showing exponential growth. Meanwhile, the per capita disposable income of rural residents has also increased year by year, from 10022.6 yuan in 2014 to 19844.8 yuan in 2023. However, common prosperity is not just about per capita income growth but also about narrowing the income gap, and while e-commerce broadens farmers’ income channels and empowers rural economic growth, it may also bring about an imbalance in the distribution of digital dividends.
The number of Taobao villages and the per capita disposable income
On the one hand, e-commerce eases information asymmetry and facilitates low-income groups to overcome the difficulty of connecting to big markets. On the other hand, farmers with higher education and better economic conditions tend to have higher digital literacy and skills so that they can enjoy more digital dividend spillover, and this “digital divide” will widen the income gap. So, in economic practice, the development of e-commerce platforms on the income level of farmers and income disparity, and then on the common prosperity of farmers and rural areas, is worth an in-depth exploration of the reality of the problem, but also the focus of this paper.
In this paper, we refer to the current common methods of academics regarding income impact analysis, as well as the e-commerce platform economy, as an explanatory variable of the relevant research to construct the basic econometric model. This paper uses the multi-period double-difference method to empirically test the impact and influence mechanism of the e-commerce platform on the common wealth of the county [23], as shown in equation (3):
Based on data availability, the Tel index and the urban-rural income ratio (urban per capita disposable income/rural per capita disposable income) were used to measure the level of regional common wealth [24]. The specific variables are selected as follows:
Explained variables Urban-rural income gap (Theil). Theil coefficient is used in the main regression to measure the income gap, and the ratio of disposable income of urban and rural residents is used for the robustness test. The Theil coefficient provides a more scientific measure by taking into account demographic changes and splitting the urban-rural income gap into between-group and within-group gaps. The larger the Thiel index, the greater the urban-rural income gap, is calculated as follows:
Core explanatory variables The level of economic development of the e-commerce platform, using the comprehensive index of economic development of the e-commerce platform (plat_), including three dimensions of the development foundation of the e-commerce platform, the level of application and the effect of the application of the eleven indicators, using entropy weighting method for calculation. Control variables In order to exclude the influence of interfering factors and ensure the robustness of the empirical results, this paper combines the existing research and selects five control variables, namely, the level of economic development (Lngdp), foreign direct investment (Lninv), government financial expenditure (Lngovproductment), the proportion of primary industry (Ind1rate) and the proportion of secondary industry (Ind2rate). Level of economic development (Lngdp). The level of economic development can play an important role in the income gap between residents of urban and rural areas. For the measurement of this indicator, this paper uses the logarithm of the per capita GDP of each city. For example, the level of economic development is an important factor affecting the urban-rural income gap. Some studies have shown that along with the increase in the level of economic development, the distribution of urban and rural industrial structures will be more reasonable, which may reduce the gap between urban and rural income levels. And the relationship between the two may change at different stages of economic development. At the early stage of economic development, some people may become rich first, and social resources may be occupied by few people, leading to the widening of the gap between urban and rural income levels. As the economy develops further, the group of people who get rich first creates more employment opportunities for society, and together with the strengthening of the government’s redistributive power, the urban-rural income gap may appear to be narrowed. Level of foreign direct investment (Lninv). The impact of foreign investment on the income gap between urban and rural residents is also multi-faceted. For example, foreign direct investment affects the flow of labor, which in turn has an impact on the urban-rural income gap. In addition, foreign direct investment is also able to affect the urban-rural income gap and the common wealth through changes in the industrial structure, employment structure, and other ways. Measured as:
Level of government expenditure (Lngovproductment). As an invisible hand, the Government can promote the redistribution of resources. In recent years, in order to develop the rural economy and improve the unbalanced development between urban and rural areas, the government has made a series of attempts to combat poverty and revitalise the countryside, and through subsidies and skills training, in order to reduce the asymmetrical development between urban and rural areas. Fiscal expenditure is a visual representation of the importance of the government in the development of each province and has a significant impact on residents’ income. The formula is:
The proportion of primary and secondary industries (Ind1rate & Ind2rate). With different weights in the industrial structure, there are differences in the distribution of labour and benefits. In the agricultural era, everyone was engaged in crop cultivation, and the distribution of income was more balanced. In the industrial era, the existence of capital prompted the industrial sector to obtain high productivity, and the collection of social wealth. The wage level is higher than in the agricultural sector, which widens the income gap between individuals and classes. It can be seen that we cannot ignore the adjustment of the industrial structure to the urban-rural income gap. The calculation formula is:
Based on the consistency and availability of data, this paper takes the 2014-2023 data of 204 prefecture-level cities (including municipalities directly under the central government) in China as the research sample, with a total of 1,950 sample observations.
Among them, the raw data for e-commerce platform development measures, urban and rural per capita disposable income data, and control variables data are all from the China Urban Statistical Yearbook.
Before conducting the empirical analysis, descriptive analyses of the variables need to be conducted, and Table 1 shows the descriptive statistics of the variables. The mean value of the explanatory variable Theil’s index (Theil) is 0.083, the maximum and minimum values are 0.278 and -0.360, respectively, and the standard deviation is 0.041. The mean value of the urban-rural disposable income ratio (Gap) is 2.383, the maximum value is 4.630, the minimum value is 1.490, and the standard deviation is 0.485. The mean value of the development of an e-commerce platform is 0.073±0.057. In addition, the mean value of the proportion of primary industry amount secondary industry is 12.258% and 48.264%, respectively, which shows that there is a large gap in the industrial structure of each city. The variables are all relatively smooth, and no individual extreme values exist.
Descriptive statistics of the variables
| Varname | Obs | Mean | SD | Min | Median | Max |
|---|---|---|---|---|---|---|
| Theil | 1950 | 0.083 | 0.041 | -0.36033 | 0.07244 | 0.27845 |
| Gap | 1950 | 2.383 | 0.485 | 1.48987 | 2.31332 | 4.62991 |
| Plat_ | 2678 | 0.073 | 0.057 | 0.0065 | 0.06473 | 0.73302 |
| lngdp | 2000 | 10.631 | 0.584 | 8.77108 | 10.59723 | 13.05069 |
| lninv | 1888 | 10.156 | 1.779 | 1.37829 | 10.19045 | 14.94327 |
| lngovprocument | 2254 | 0.108 | 0.094 | 0.01023 | 0.07665 | 1.26686 |
| Lnd1rate | 2020 | 12.258 | 7.725 | 0.031 | 11.38 | 49.884 |
| Lnd2rate | 2020 | 48.264 | 10.38 | 14.948 | 48.543 | 89.336 |
Before the regression analysis, the LLC unit root check method is chosen to carry out the smoothness test, and the test result rejects the original hypothesis of the existence of a unit root, which indicates that the data are smooth and avoids the occurrence of pseudo-regression problems due to the non-smoothness of the data. Next, it is necessary to use the Variance Inflation Factor (VIF) to test whether there is a serious covariance between the variables. The results of the VIF test of each variable are shown in Table 2, and it is found that the Mean VIF is 1.60, and the value of the VIF of each variable is less than 2, which indicates that there is no multicollinearity between the variables.
Results of the VIF test for each variable
| Varname | VIF | 1/VIF |
|---|---|---|
| Theil | 1.96 | 0.51 |
| Plat_ | 1.71 | 0.58 |
| lngdp | 1.67 | 0.60 |
| lninv | 1.62 | 0.62 |
| lngovprocument | 1.55 | 0.65 |
| Lnd1rate | 1.43 | 0.70 |
| Lnd2rate | 1.25 | 0.80 |
| Mean VIF | 1.60 | |
In order to test the impact of e-commerce platform development on regional common wealth, this paper estimates equation (1), and the results of the base regression are shown in Table 3, *** p<0.01, **p<0.05, *p<0.1, and robust standard errors in parentheses, which are the same as in the following table. Column (1) presents the results of the base regression, column (2) presents the results after adding all control variables, and column (3) reports the results of the regression with both year and region fixed effects on top of the column (2), respectively. The results show that with respect to the core variables, the quality of e-commerce platform development on the urban-rural income gap (Theil) is significantly negative at the 1 per cent level in all regressions, indicating that the development of e-commerce platforms significantly reduces the urban-rural income gap, and more specifically, every 1 per cent of the development of e-commerce platforms will contribute to an average increase in the level of regional common wealth by 0.011 per cent per year. This proves that the development of e-commerce platforms has a positive effect on narrowing the urban-rural income gap and plays the role of “more good, more good” in the process of promoting regional common wealth.
Benchmark regression results
| (1) | (2) | (3) | |
|---|---|---|---|
| Plat_ | -0.025***(0.028) | -0.017***(0.005) | -0.011***(0.001) |
| lngdp | -0.023***(0.008) | -0.021***(0.006) | |
| lninv | -0.019***(0.076) | -0.015***(0.042) | |
| lngovprocument | -0.021***(0.052) | -0.015***(0.038) | |
| Lnd1rate | -0.002*(0.018) | -0.011**(0.021) | |
| Lnd2rate | -0.001*(0.004) | -0.005*(0.001) | |
| -10.259***(0.152) | -10.018***(0.002) | -10.418***(0.037) | |
| Region | NO | NO | YES |
| Year | NO | NO | YES |
| Observed | 1950 | 1950 | 1950 |
| 0.885 | 0.912 | 0.976 |
The use of a multi-period DID requires that the demonstration regions (regions with excellent development of the e-commerce platform) and non-demonstration regions are not significantly different or have the same trend on urban-rural income inequality before the optimisation of the e-commerce platform data-driven strategy. Figure 2 shows the results of the parallel trend with a 90% confidence interval. By examining the time-dynamic effects of the parallel trend and the policy, it is found that before the establishment of the demonstration regions, none of their estimates of urban-rural income inequality were significant, and the coefficient value changes more gently and tends to zero, so it can be seen that the assumption of the parallel trend is valid.
Parallel trend test results
In addition, this paper utilizes the method of randomly generating the e-commerce platform into the integrated demonstration area pilot strategy time and area as a placebo test. Figure 3 shows the distribution of the estimated coefficients of the randomly generated e-commerce platform into the demonstration region strategy repeated 2000 times. It can be found that the estimated coefficients of the placebo test are mainly concentrated around 0, significantly different from the estimated coefficients under the real e-commerce platform demonstration region is a data-driven strategy, and most of the estimation results are not significant, which means that the omitted variables have less impact on the estimation results, and the estimation results of this paper are more robust.
Placebo test knot
The sample data are divided into two regions, the East and the Midwest, to analyze the heterogeneity of the impact of different regional characteristics of e-commerce platforms on urban-rural income differences. Table 4 shows the analysis of income heterogeneity between urban and rural areas by region. The regression coefficients of e-commerce platform strategy optimisation for high urban-rural income differences in eastern China are significantly negative at the 5% level, while the impact on low urban-rural income differences in eastern China is not significant. The regression coefficient of the e-commerce platform for high urban-rural income differences in central and western China is significantly positive at the 10% level, while the effect on low urban-rural income differences in central and western China is not significant. In the era of rapid development of e-commerce platforms, due to the vastness of China’s territory, the phenomenon of regional development imbalance is prominent. The central and western provinces have a certain “innate” positional disadvantage, which makes the region’s enterprises and units can not enjoy the dividends of the development of e-commerce platform, and a part of the employees rely on a fixed salary for a living will choose to leave their units and choose to engage in agriculture and forestry. Choose to leave the unit, choose to engage in agriculture, forestry, animal husbandry and fishery activities to make a living in order to make better use of the unique geographic advantages of the central and western regions, which makes the central and western regions of the urban-rural income disparity.
Analysis of the heterogeneity of urban and rural income by region
| East | Central and Western | |||
|---|---|---|---|---|
| Low | High | Low | High | |
| Plat_ | -0.068(0.992) | -0.151**(2.256) | 0.155(1.422) | 0.1525*(-1.722) |
| Control | Control | Control | Control | Control |
| Constant | Control | Control | Control | Control |
| Region | Control | Control | Control | Control |
| Year | Control | Control | Control | Control |
| Observed | 880 | 1070 | 968 | 982 |
| 0.3678 | 0.5781 | 0.3412 | 0.5561 | |
Continuously expanding inequality in farm household incomes can hinder the process of common prosperity, so policymakers are focused on effectively bridging the inequality caused by the digital economy represented by e-commerce platforms. This section attempts to build three external environments, namely, traditional infrastructure construction, digital infrastructure construction and regional human capital elite, to study the heterogeneous impacts of the improvement of external environmental constraints on different social groups in order to explore how to bridge the rural income inequality amplified by the development of e-commerce platforms through the unification of external pull and endogenous dynamics, so as to build a feasible channel for low-income regions to share the fruits of the economy and realise the common wealth of the region. Feasible channels. Table 5 shows the results of the heterogeneity test of the impact of e-commerce platform strategy optimization on rural-urban income inequality.
Traditional Infrastructure Development This paper firstly analyses the heterogeneity of the impact of e-commerce platform development on urban-rural income inequality from the dimension of traditional infrastructure construction. The variable “traditional infrastructure construction” is constructed based on whether there is investment in rural roads in the database, and if there is, it is assigned the value of “1”, and vice versa, it is assigned the value of “0”. As shown in column (1), the regression coefficient of the interaction term between e-commerce platform demonstration regions and traditional infrastructure development is -0.019, which indicates that the impact of e-commerce platform development on rural-urban income inequality is significantly reduced by 1.9 percentage points in demonstration regions with traditional infrastructure investment relative to those without. The above findings suggest that improving traditional infrastructure development is an effective bridge to the widening of the urban-rural income gap through e-commerce platform development. Digital Infrastructure Development Heterogeneity in the impact of rural e-commerce on urban-rural income inequality is investigated from the perspective of digital infrastructure building.
Digital infrastructure building (household level) In this paper, we set a dummy variable for households’ Internet access, which takes the value of “1” if they have Internet access and “0” otherwise. The estimation results are shown in column (2). The interaction term between the e-commerce platform demonstration area and household Internet access is -0.018, which indicates that in the demonstration area, households with Internet access face significantly lower income inequality than those without Internet access by 1.8 percentage points. Internet access at the household level can mitigate the amplifying effect of e-commerce development on urban-rural income inequality. Digital infrastructure development (regional dimension) This paper analyses the heterogeneity of the optimisation of e-commerce platforms affecting urban-rural income inequality in terms of the dimension of digital infrastructure building at the regional level and characterises digital infrastructure building in terms of the region’s Internet penetration rate obtained from the number of households with Internet access in the region as a percentage of the region’s total number of households. To a certain extent, a higher regional Internet penetration rate means a higher level of digitisation, and this paper defines a high Internet penetration rate based on the median Internet penetration rate in the region, which takes the value of “1” if the regional Internet penetration rate is higher than the median, and “0” otherwise. The estimated coefficient of the interaction term in the result in column (3) is significantly negative, which suggests that in the process of the development of e-commerce platforms, the rise of regional Internet penetration will also effectively bridge the urban-rural income inequality caused by the development of e-commerce platforms. Regional human capital elite This paper analyses the heterogeneous impact of the development of e-commerce platforms on urban-rural income inequality in terms of regional “human capital elites”. The proportion of village cadres with an education level of high school and above is used to measure the rural “human capital elite”, and the median is used to distinguish between high and low categories. The regression results, as shown in (4), show that the coefficient of the interaction term between e-commerce demonstration regions and high rural human capital elites is -0.018 and is significant at the 5 per cent level, which suggests that, in the demonstration regions, compared with those with low rural human capital elites, those with high regional human capital elites are able to significantly and efficiently bridge urban-rural income inequality due to the development of e-commerce platforms.
The e-commerce platform strategy optimizes the heterogeneity
| Variable | Urban and rural income inequality | |||
|---|---|---|---|---|
| (1) | (2) | (3) | (4) | |
| Plat_ | 0.022***(0.005) | 0.018***(0.005) | 0.022**(0.012) | 0.021**(0.008) |
| Plat_ × Tic | -0.019***(0.005) | |||
| Plat_ × Hic | -0.018**(0.008) | |||
| Plat_ × Hipr | -0.021*(0.015) | |||
| Plat_ × Hirh | -0.018**(0.002) | |||
| Control | Control | Control | Control | Control |
| Region | Control | Control | Control | Control |
| Year | Control | Control | Control | Control |
| Observed | 1950 | 1950 | 1950 | 1950 |
| 0.244 | 0.232 | 0.244 | 0.244 | |
There are significant differences in various aspects between regions, which will naturally lead to unbalanced development. To solve the problem of unbalanced regional development of rural e-commerce, we can start from the following aspects. Formulate a clear regional development plan to clarify the development priorities and goals of each region to ensure a more balanced distribution of resources. Increase investment in infrastructure in underdeveloped regions, including roads, bridges, electricity, water resources, etc., in order to increase productivity and attract investment. Encourage and support small and microenterprise entrepreneurship, which will help to create more employment opportunities and also promote the diversification of the economic structure. Improve public services in underdeveloped regions, such as in the areas of health care, sanitation, security, and social welfare, in order to raise the living standards and sense of well-being of the population.
E-commerce can promote the development of the agricultural products market to a large extent, but in order to fully open up the online agricultural products market, it is necessary to take a series of key measures and solve a series of problems. For example, it is necessary to strengthen the construction of rural transport, communications and other infrastructures, formulate quality standards for agricultural products, strictly control the quality of products, improve the logistics and distribution network, reduce logistics costs, and enhance the capacity of cold-chain logistics and distribution.
To create product characteristics, it is necessary to conduct thorough market research, gain an in-depth understanding of market demand, and identify the correct market segments. Secondly, it is necessary to continuously increase investment in research and development, innovative product design, and functionality to meet the changing needs of consumers.
To pay attention to brand building, the brand reflects the comprehensive strength and market recognition of an enterprise. Rural e-commerce enterprises should establish brand awareness from the beginning, strive to improve user reputation, operate practically, and constantly grow their business.
To solve the problem of financing e-commerce, local governments need to consider various financing channels and strategies to meet the needs of various e-commerce enterprises. Create a special rural e-commerce fund to provide financial support and risk sharing to attract more investors. Work with local rural co-ops and credit unions to leverage their resources and networks to provide financing support. Encourage rural e-commerce enterprises to become members of co-operatives or credit unions to gain access to more financing opportunities. Promote the development of special financing products for rural eCommerce by banks and financial institutions to meet the financing needs of enterprises.
Promoting regional common prosperity is one of the key areas of focus in promoting the common prosperity of all people, and it is necessary to effectively control and continuously reduce the urban-rural income gap and achieve inclusive growth that takes into account both growth and distribution. The optimization of data-driven strategies on e-commerce platforms has promoted the transformation of economic and social development, facilitated the effective connection between farmers and the market, led to the growth of employment and entrepreneurship among farmers, and significantly weakened the information disadvantage of low-income farmers, thereby empowering farmers and rural communities to share prosperity by reducing the urban-rural income gap.
This paper uses a two-way fixed-effects model to look at data from 204 prefecture-level cities in China from 2014–2023. These cities include municipalities directly under the central government. The study looks at how e-commerce platform data-driven strategy optimization affects regional common wealth. The results found that e-commerce platform data-driven strategy optimization significantly promotes regional common wealth and suppresses urban and rural disposable income per capita, and this finding still holds after a series of robustness tests. Heterogeneity analyses reveal that regions with high levels of traditional infrastructure development, digital infrastructure development, and village cadres’ human capital experience a reduction in the amplifying effect of e-commerce development on farm household income inequality.
E-commerce platform strategies to optimize the level of shared prosperity do not work through a single factor but rather are the result of a combination of factors. Create a unique rural e-commerce fund that offers financial assistance and risk sharing to entice additional investors. Work with local rural co-operatives and credit unions to leverage their resources and networks to provide financing support. Encourage rural e-commerce enterprises to become members of cooperatives or credit unions to gain access to more financing opportunities. Promote the development of special rural e-commerce financing products by banks and financial institutions to meet the financing needs of enterprises.