Introduction

As a national policy issued in 2013, the Broadband China (BC) policy focuses on facilitating the rapid development of internet infrastructure. Its objective is to elevate the broadband network infrastructure to a level comparable to that of developed nations by 2020 (Peng et al., 2023; Zuo et al., 2024). Furthermore, the BC policy emphasizes the imperative of establishing broadband networks. It specifies that the purpose is to achieve a fixed broadband household penetration rate exceeding 70%, with particular emphasis on coverage of over 98% in rural areas. Consequently, this strategic initiative represents a pivotal component of the Chinese government’s comprehensive information infrastructure construction plan, which exerts a significant influence on the economy, especially in rural areas.

The income gap is a crucial concern for the economic development trajectory of every nation. Moreover, poverty eradication stands as the foremost theme within the United Nations’ 2030 Sustainable Development Goals (Tang et al., 2022). In China, significant strides have been made in alleviating poverty, but measures are required to address the widening gap between urban and rural incomes (Zhang, 2021; Gao and Li, 2023; He et al., 2024). Coincidentally, the aim of BC policy is to rectify imbalances in urban-rural development by promoting broadband infrastructure across China. Therefore, an examination of the correlation between BC policy and urban-rural income disparity holds momentous practical significance. It may also enable other countries to glean developmental insights from Chinese policies.

In 2013, the government announced the “Broadband China” strategy and a specific implementation planFootnote 1. This strategy explains in detail the goals, methods, and stages of information infrastructure construction. It emphasizes addressing the problems of broadband network construction due to unclear positioning of infrastructure, development imbalances, and insufficient original technology capabilities. Furthermore, the BC policy aims to narrow the gap between broadband infrastructure development and that of developed countries by 2020, ensuring that Chinese citizens can fully benefit from economic growth, enhanced service accessibility, and development opportunities facilitated by broadband technology. In other words, with the implementation of the BC policy, the broadband is poised to catalyze nationwide network popularization. At the same time, universal access to the internet is anticipated to create more income-earning opportunities for both urban and rural residents (Zhou et al., 2020). Besides, the advent of information technology has reduced barriers to employment across various industries and had a substantial impact on income distribution (Su et al., 2022; Das and Chatterjee, 2023). In summary, the BC policy will have a significant impact on the income of urban and rural residents in China.

The Chinese Ministry of Industry and Information Technology released the first batch of Demonstration Cities lists in 2014 to effectively implement the requirements of the BC strategyFootnote 2. In other words, the cities in this list will prioritize the acceleration of information infrastructure construction based on the BC Strategy. Therefore, applying for the demonstration cities needs a stronger foundation in information technology and a clearer construction project. After that, China’s information sector will provide priority support to the demonstration cities for new information technology, operations, and other pilot projects. Aside from that, the demonstration cities will implement measures to integrate broadband network construction into local urban-rural planning. This implies that the cities will continue to advance in terms of networking and user-level development. After 2014, China established two more demonstration cities in 2015 and 2016. Due to the pilot program only selecting partial areas within certain cities, it is not feasible to assess its impact at the city-level. Consequently, this study excluded such cities from the overall sample; for instance, only Jiangjin District and Rongchang District in Chongqing were designated as pilot cities in 2015. At the same time, considering the availability of the data, this study selected 94 cities from all the 117 sample cities as treatment groups. Figure 1 displays the distribution of the cities in the treatment groups.

Fig. 1
figure 1

Distribution of the cities in the treatment groups.

Full size image

The pilot cities offers an opportunity to investigate the effects of the Broadband China (BC) policy. In previous studies, pilot cities are often employed as dummy variables to explore the alterations brought about by the BC policy. For instance, the BC policy has exerted a positive influence on income, green innovation, and high-quality economic development (Fang et al., 2022; Das and Chatterjee, 2023; Hong et al., 2023). Nevertheless, when it comes to the balance of urban and rural development, the existing literature is scarce. Although some scholars have deliberated on the digital economy and the urban-rural income gap, relatively few have initiated from the policy level (Deng et al., 2023; Dong et al., 2023; Gao and Li, 2023). Specifically, there is no profound discussion on policy coordination and the analysis of the influence mechanism. With reference to the relevant policy documents mentioned above, alleviating the urban-rural gap will be one of the significant development directions of the BC policy. Therefore, the Broadband China policy should be conducted in-depth on this aspect.

Although the BC policy enhances the incomes of both urban and rural residents, urban areas, due to their stronger economic foundation, may obtain more dividends compared to rural areas (Nie and Wan, 2023). Therefore, this study primarily investigates the influence of the BC policy on the urban-rural income gap and conducts more profound research in comparison with the extant literature. It offers the following possible contributions: Firstly, it examines the relevant BC policy documents and conducts further research using the guidance provided by the BC strategy. Secondly, it is committed to enriching the mechanism by which the BC policy influences the urban-rural income gap. Last but not least, this study investigates whether the BC policy can be combined with other related policies to generate synergies.

This paper is structured in following manners: Section two is the literature review and theoretical hypotheses. Section three is the empirical model and data description, and section four presents the results of the empirical analysis. In the end, Section five presents the conclusions, implications and recommendations.

Literature review and theoretical hypotheses

Literature review

In recent years, with the expansion of the digital economy, the relationship between the digital economy and income disparities has been investigated. In the study of 49 countries, researchers found that the digital economy plays an indirect role in reducing income inequality by fostering advancements in industrial structure (Xu, 2023). But this effect is not obvious in countries with large income gaps. Das and Chatterjee (2023) found that the spread of communication technologies in India improved persistent poverty, where digital finance had a positive impact on income inequalities. However, Peng and Dan (2023) discovered that the digital economy exacerbated income disparities, revealing a U-shaped relationship between them. The impact of information technology on income inequality is commonly intertwined with other technological and political factors (Bauer, 2017). Based on the above, the impact of digitalization on economics is not only upon the digital infrastructure but also on local policies. The development of broadband infrastructure in China will inevitably lead to an increase in residents’ income. Due to various factors, the income growth of urban residents may have been more significant than that of rural residents (Nie and Wan, 2023). Hence, it is necessary to have more in-depth studies on the relationship between BC policy and urban-rural income inequality.

The perspective of infrastructure construction is one of the crucial directions for investigating the relationship between the digital economy and the income gap. According to Houngbonon and Liang (2021), the expansion of broadband internet in France reduced local income inequality by increasing employment in manual jobs. The study by Zolfaghari et al. (2020) revealed that investment in communication technology infrastructures had a great impact on reducing income inequality in Iran. Meanwhile, broadband internet provides a digital dividend for the low-income brackets in China, and it favors highly professional workers. Besides, Wang and He (2024) found that the expansion of broadband infrastructure can assist rural households in managing unforeseen health crises and decreasing the risk of slipping into poverty. Similarly, the development of digital infrastructure is increasingly impacting household income and income inequality (Lu et al., 2023). In summary, “Broadband China” functions as a valuable quasi-natural experiment, which makes it of great significance to conduct an in-depth analysis of its impact on the income gap.

Research on BC policy is expanding in scope, with a focus on employment dynamics, digital transformation, innovation initiatives, and so on. Considering BC policy as a quasi-natural experiment, Fang et al. (2022) discovered that internet development enhances innovation efficiency. In the same way, Hong et al. (2023) found that network infrastructure construction reduced energy intensity through the mediating effects of energy efficiency and innovation. Moreover, digital transformation, through technological optimization and industrial upgrading, enables BC policy to reduce electricity consumption and intensity (Wang et al., 2022). More importantly, the study by Zhou et al. (2022) showed that BC policy influences export trade growth by improving information efficiency directly, leading to lower logistics costs, enhanced trade efficiency, and reduced barriers to trade.

In addition, there is little research on the relationship between the “Broadband China” policy and the income disparity. Deng et al. (2023) provided a brief discussion about how BC policy have contributed to the widening income gap. Nie and Wan (2023) illustrated that the expansion of the income gap is contingent upon educational attainment and exhibits variability across diverse geographical areas. Other literature, focusing on different aspects, had demonstrated the impact of the BC policy on industrial structure, employment, innovation, the market environment, and export trade (Wang et al., 2022; Zhou et al., 2022; Jia et al., 2023; Du et al., 2024; Wei et al., 2024). Moreover, the documents concerning the BC policy also emphasized the significance of constructing broadband networks that prioritize application capacity, service level, and independent innovations to ensure the full benefit of nationals from economic growth and development opportunities. Therefore, it’s important to further examine how BC policy influences the income gap while considering economic factors.

Comparatively, this study offers the following contributions: First, it contributes to enriching the mechanism through which BC policy influences the urban-rural income gap. The past literature reveals that BC policy may influence factors such as innovation and entrepreneurship, which serve as crucial mediating variables. Second, this study investigates the potential synergistic effects of other policies. Over the past decade, the Chinese government has implemented various national policies, such as “the National E-commerce Demonstration Cities Policy”, “the National Information Consumption Pilot Policy”, and “the Technological Finance Cooperation Pilot Policy”, etc. Furthermore, the implementation of these policies frequently intersects with BC policy, potentially influencing its effectiveness. As a result, it is essential to analyze the impact of BC policy from a policy coordination perspective. Finally, this paper thoroughly studies the relevant BC policy documents and conducts further research in alignment with the guidance provided by BC strategy.

Theoretical hypotheses

The Broadband China demonstration city construction is one of the main programs to promote the Chinese digital infrastructure under the BC strategy. First, the BC demonstration cities have inherent advantages in terms of digital facilities, which provide essential support for internet facility construction. Second, BC demonstration cities receive policy support and economic assistance in the related aspects of broadband construction. This means that these pilot cities have an economic foundation and development potential that non-pilot cities lack. Therefore, using the BC demonstration cities as the treatment group in this study is reasonable. Besides, in the early days of the widespread use of information and communications technology (ICT) in rural areas, ICT had a significant impact on decreasing the Chinese urban-rural income gap. In particular, the use of smartphones has significantly increased income from agricultural products in rural areas (Dong et al., 2023). Nowadays, with the construction of BC demonstration cities underway, there is a rapid development of network infrastructure in urban and rural areas, and it has a strong impact on the income gap.

Commonly, there are two potential pathways through which BC policy influences the income gap. First, the “Broadband China” policy significantly reduces the urban-rural income disparity. This is because the construction of broadband networks has lowered network costs for farmers and increased operational efficiency. Hence, the earning potential in rural areas has been fully realized, leading to a rapid convergence of rural and urban incomes. Second, digital infrastructure significantly broadens the income gap. As a result of a stronger economic base, urban areas have better conditions to absorb the benefits of the BC policy and gain more economic dividends than rural areas. This study believes that all the residents can benefit from the BC policy, while the urban residents may gain more welfare, and the income gap may divide. Hence, this study proposes the hypotheses H1.

H1: The Broadband China policy widens the urban and rural income gap.

With the rapid development of the digital economy, new industries and products have emerged globally. Subsequently, the integration of the real world has brought applications in various aspects of life and work, such as artificial intelligence, smart homes, e-commerce, etc. Existing literature has identified self-enhancing innovation mechanisms within information infrastructure that can generate and provide new services (Bygstad, 2010). Coincidentally, these new services can often address urgent needs within local communities. For example, from 2002 to 2017, China’s information infrastructure facilitated innovation through improved channels of exchange between different enterprises (Lumeng et al., 2023).

The BC policy has been a crucial framework for information infrastructure development in China for the past decade. It prioritizes the integration of broadband network construction with independent innovation, aiming to stimulate innovation and entrepreneurship in the pilot cities. When it comes to innovation, participants can achieve innovative outcomes or increase their earnings. Furthermore, entrepreneurship can stimulate industrial advancement, boost workforce earnings, and contribute to job creation. On the one hand, broadband infrastructure in rural areas presents opportunities for digital agriculture, unmanned aerial vehicle technology, distance education, and rural e-commerce. Furthermore, the advancement of digital finance offers convenience for rural entrepreneurship, facilitating the development of rural products and tourism beyond geographical limitations. Simultaneously, broadband connectivity enables access to external cultural resources, fostering innovative development within rural culture. On the other hand, broadband deployment accelerates advancements of information technology in urban areas, such as software development and big data analytics. This facilitates more efficient product innovation and industrial development within companies. Additionally, the availability of internet connectivity has created new entrepreneurial opportunities through cross-industry integration and diversification of consumer needs. In conclusion, it is likely to have an influence on local employment and income (Wang et al., 2023). Similarly, studies have illuminated that innovation and entrepreneurship can influence income inequality (Law et al., 2020; Hu et al., 2023; Ongo et al., 2024). Therefore, this study hypothesizes that the BC policy influences the income gap through innovation and entrepreneurship, denoted as H2.

H2: The Broadband China policy affects the urban-rural income gap by influencing innovation and entrepreneurship.

The coordinated development of network construction, application popularization, and industrial support provide important guidance for the implementation of the BC policy. Consequently, digital industries are bound to receive support from relevant government departments and enterprises. Particularly, the digital finance sector and information service industry are poised for inevitable growth, driven by infrastructure development and policy dividends. Meanwhile, innovation, investments, non-farm employment, etc., influenced by digital inclusive finance, will increase the income of rural residents (Lian et al., 2023). Similarly, development in the information industry will reduce network costs, increase application services, and ultimately affect local employment and income. Therefore, BC policy affects the income gap by promoting the development of digital inclusive finance and information industries in demonstration cities, and this is recorded as hypothesis H3.

H3: Digital inclusive finance and information sectors are the mechanisms by which the Broadband China policy affects the urban-rural income gap.

Apart from the Broadband China (BC) policy, the government has also implemented some other policies aimed at promoting the development of digital industries. Interestingly, the policy document of the BC demonstration cities emphasizes priority support for these cities in pilot projects of new information and communication technologies as well as in services. As a result, there are overlaps between the BC demonstration cities and several other pilot cities, which implies that a comprehensive assessment of the impact of other policies is also necessary when analyzing the BC policy.

As the BC policy is focused on the construction of broadband networks, its effect is influenced by industry, consumption, and technology. This study selects three pilot policies as the focus to examine their synergistic effects with BC policy. First, the “National E-commerce Demonstration Cities Policy”. These cities leverage e-commerce as the foundation for optimizing resource allocation and industrial structure. Second, the “National Information Consumption Pilot Cities Policy”. To address the challenges in information consumption, such as limited product and service innovation and significant industry obstacles, Chinese governments have implemented this policy. Meanwhile, it stimulates information consumption and facilitates the upgrading of consumption and industrial transformation. Third, the “Technological Finance Cooperation Pilot Policy”. This policy aims to innovate financial products and establish service platforms through innovative financial investment. Its fundamental objective is to achieve the seamless integration of the innovation chain with the capital chain in finance while offering tailored financial services for science and technology enterprises at every stage of their development. What’s more, one of its goals is to promote the commercialization of technological breakthroughs, strengthen capacity for innovation, and nurture strategic emerging industries, which are their primary focus areas.

These policies may have impacts on the economic role of broadband networks. In past studies, Liu et al. (2024) discovered that digitization has a significant impact on residents’ consumption and income by influencing e-commerce. Interestingly, the objective of the “National E-commerce Demonstration Cities Policy” is to advance the growth of e-commerce, thereby emphasizing the significance of digital infrastructure. Similarly, the “National Information Consumption Pilot Cities Policy” can drive the development of the information service industry and stimulate residents’ consumption. It means that the broadband construction promoted by the BC policy will play a greater role. Furthermore, the “Technological Finance Cooperation Pilot Policy” can address several issues, including inadequate support for innovation, deficiency in technologies, and environmental degradation (He et al., 2024). The impact aligns with the principles emphasized in BC policy, which mandate integrating infrastructure enhancement with application services and industrial innovation. In other words, these policies may influence the impact of BC policy, and hence hypothesis H4 is proposed.

H4: The Broadband China policy generates synergistic impacts when combined with other policies.

Empirical model and data description

Empirical model

By conducting a comparative analysis of the treatment group and the control group before and after the policy, the difference-in-differences (DID) model effectively measures the impact of the policy. Moreover, the utilization of the DID model may partially alleviate the problem of endogeneity arising from sample selection bias (Li et al., 2021). In different years, BC policy has established three batches of pilot cities. According to Beck et al. (2010) and Ding et al. (2024), employing a multi-period DID approach can provide a more precise measurement of the policy’s effect with different time. To investigate the influence of the BC policy on the income gap, this study utilizes the multi-period DID approach for analysis. According to the list of “Broadband China” demonstration cities and considering the data availability, 94 cities and 151 cities are involved in the treatment and control groups, respectively. Equation (1) presents the model of DID analysis:

$${{{{\mathbf{GAP}}}}}_{{{{\mathbf{it}}}}}={{\boldsymbol{\alpha }}}_{{\bf{0}}}+{{\boldsymbol{\alpha }}}_{{\bf{1}}}{{{{\mathbf{BC}}}}}_{{{{\mathbf{it}}}}}+{{\boldsymbol{\alpha }}}_{{\bf{2}}}{{{{\mathbf{Control}}}}}_{{{{\mathbf{it}}}}}+{{\boldsymbol{\phi }}}_{{\boldsymbol{i}}}+{{\boldsymbol{\varphi }}}_{{\boldsymbol{t}}}+{{\boldsymbol{\varepsilon }}}_{{{{\mathbf{it}}}}}$$
(1)

\({{{{\mathbf{GAP}}}}}_{{{{\mathbf{it}}}}}\) is measured by the Theil index, which shows the urban-rural income gap in city \({\boldsymbol{i}}\) at year \({\boldsymbol{t}}\). To enhance the clarity of the regression results, the Theil index is multiplied by 100. \({{{{\mathbf{BC}}}}}_{{\boldsymbol{it}}}\) is dummy variable. If the city i is included in BC policy in year \({\boldsymbol{t}}\), \({{{{\mathbf{BC}}}}}_{{\boldsymbol{it}}}\) equals 1, otherwise 0. Besides, \({{{{\mathbf{Control}}}}}_{{\boldsymbol{it}}}\) represents the control variables.

To test the mechanisms of the BC policy affecting the income gap, a gradual regression model is constructed based on Eq. (1). The specific equations are as follows:

$${{{{\mathbf{Mediator}}}}}_{{{{\mathbf{it}}}}}={{\boldsymbol{\alpha }}}_{{\bf{0}}}+{{\boldsymbol{\alpha }}}_{{\bf{1}}}{{{{\mathbf{BC}}}}}_{{\boldsymbol{i}}{\boldsymbol{t}}}{+{\boldsymbol{\alpha }}}_{{\bf{2}}}{{{{\mathbf{Control}}}}}_{{{{\mathbf{it}}}}}+{{\boldsymbol{\phi }}}_{{\boldsymbol{i}}}+{{\boldsymbol{\varphi }}}_{{\boldsymbol{t}}}+{{\boldsymbol{\varepsilon }}}_{{{{\mathbf{it}}}}}$$
(2)
$${{{{\mathbf{Gap}}}}}_{{{{\mathbf{it}}}}}={{\boldsymbol{\alpha }}}_{{\bf{0}}}+{{\boldsymbol{\alpha }}}_{{\bf{1}}}{{{{\mathbf{BC}}}}}_{{\boldsymbol{i}}{\boldsymbol{t}}}+{{\boldsymbol{\alpha }}}_{{\bf{2}}}{{{{\mathbf{Mediator}}}}}_{{\boldsymbol{i}}{\boldsymbol{t}}}{+{\boldsymbol{\alpha }}}_{{\bf{3}}}{{{{\mathbf{Control}}}}}_{{{{\mathbf{it}}}}}+{{\boldsymbol{\phi }}}_{{\boldsymbol{i}}}+{{\boldsymbol{\varphi }}}_{{\boldsymbol{t}}}+{{\boldsymbol{\varepsilon }}}_{{{{\mathbf{it}}}}}$$
(3)

\({{\boldsymbol{Mediator}}}_{{\boldsymbol{it}}}\) represents the mediator variables. If the \({{\boldsymbol{\alpha }}}_{{\bf{1}}}\) of Eq. (1) and Eq. (2) are both significant and the \({{\boldsymbol{\alpha }}}_{{\bf{2}}}\) of Eq. (3) is significant, it indicates the existence of a mediating effect (Yu et al., 2021; Zhang et al., 2024).

To test the synergistic effect between the BC policy and other policies, a model has been further constructed based on Eq. (1). As shown in the Eq. (4) below:

$${{{{\mathbf{Gap}}}}}_{{{{\mathbf{it}}}}}={{\boldsymbol{\alpha }}}_{{\bf{0}}}+{{\boldsymbol{\alpha }}}_{{\bf{1}}}{{{{\mathbf{BC}}}}}_{{\boldsymbol{i}}{\boldsymbol{t}}}+{{\boldsymbol{\alpha }}}_{{\bf{2}}}{{{{\mathbf{P}}}}{{{\mathbf{OLICY}}}}}_{{\boldsymbol{i}}{\boldsymbol{t}}}{+{{\boldsymbol{\alpha }}}_{{\bf{3}}}{{{{\mathbf{BC}}}}{\rm{\_}}{{{\mathbf{POLICY}}}}}_{{{{\mathbf{it}}}}}+{\boldsymbol{\alpha }}}_{{\bf{4}}}{{{{\mathbf{Control}}}}}_{{{{\mathbf{it}}}}}+{{\boldsymbol{\phi }}}_{{\boldsymbol{i}}}+{{\boldsymbol{\varphi }}}_{{\boldsymbol{t}}}+{{\boldsymbol{\varepsilon }}}_{{{{\mathbf{it}}}}}$$
(4)

\({{{{\mathbf{POLICY}}}}}_{{{{\mathbf{it}}}}}\,\) represent the dummy variables of other policy, it equals 1 if city \({\boldsymbol{i}}\) is a polit city of a policy in year \({\boldsymbol{t}}\), otherwise 0. \({{{{\mathbf{BC}}}}{\rm{\_}}{{{\mathbf{POLICY}}}}}_{{{{\mathbf{it}}}}}\) is the interaction term of \({{{{\mathbf{BC}}}}}_{{{{\mathbf{it}}}}}\) and \({{{{\mathbf{POLICY}}}}}_{{{{\mathbf{it}}}}}\). \({{\boldsymbol{\alpha }}}_{{\bf{3}}}\) in Eq. (4) reflects the synergistic effect of other policies and BC policy. If it is significantly positive, it indicates that this policy enhanced the effectiveness of BC policy.

Data description

The Thiel Index is an indicator that captures the dynamics of the urban-rural income gap. It not only illustrates the income disparity between urban and rural areas but also takes population factors into account. Hence, plenty of studies have employed the Thiel Index as a measure of the income gap (Dong and Hao, 2018; Tang et al., 2022), and its calculation is shown in Eq. (4) below:

$${{\mathrm{T}}{{\mathrm{heil}}}}_{t}=\displaystyle\mathop{\sum}\limits_{n=1}^{2}\frac{{I}_{{nt}}}{{I}_{t}}\mathrm{ln}\left(\frac{{I}_{nt}/{P}_{nt}}{{I}_{t}/{P}_{t}}\right)$$
(5)

\({{Theil}}_{t}\) represents the Theil index in year t. \({I}_{1t}\) and \({I}_{2t}\) represent urban total income and rural total income, respectively, in year t. \({P}_{1t}\) and \({P}_{2t}\) represent the urban total population and rural total population, respectively, in year t. \({I}_{t}\,\) equals the sum of \({I}_{1t}\) and \({I}_{2t}\). \({P}_{t}\,\) equals the sum of \({P}_{1t}\) and \({P}_{2t}\). The Thiel index is positively correlated with the income gap.

There are numerous factors that influence the income gap. To minimize the result errors, it is essential to control for significant interfering variables, and the following variables are used as control variables (Su et al., 2015; Tang et al., 2022; Xia et al., 2024). The calculation method of control variables is referred to in the above literature. The urbanization rate (UTR) is represented by the proportion of the total urban population to the total population at the end of the year; The level of foreign investment (FDI), represented by the actual utilization of foreign direct investment as a proportion of GDP; The level of human capital (HUMAN), represented by the ratio between the number of students in the general specialty and the total population at the end of the year; real GDP (RGDP), represented by the natural logarithm of fixed GDP; The level of transportation facilities (ROAD), represented by the natural logarithm of per capita road area; The industrial structure level (IST), measured by the ratio of added value from the tertiary industry to GDP.

Based on the hypotheses, innovation, entrepreneurship, digital inclusive finance, and information sectors are possible mechanisms for the BC policy to affect the income gap. Before the analysis of hypotheses H2 and H3, it is necessary to quantify the factors with past studies (Lumeng et al., 2023; Hu et al., 2023; Ongo et al., 2024). To test the detailed mechanisms, mediator variables are selected as follows: Innovation level (INN), expressed by the number of patent applications; entrepreneurship level (ENT), represented by the number of start-ups per hundred people; Information industry development level (IT), indicated by the total number of postal and telecommunications business volumes; Digital inclusive finance (DIF), represented by the “Peking University Digital Financial Inclusion Index of China”. To examine the synergistic impact of the BC policy with other policies, the variables are included: the “National E-commerce Demonstration Cities Policy” (ECOMMERCE); the “National Information Consumption Pilot Cities Policy” (CONSUMPTION); the “Technological Finance Cooperation Pilot Policy” (TECHFIANCE). These policies are represented by the dummy variables, which equals 1 if the city is designated as a pilot city, otherwise 0.

To enhance the integrity of the data, this study opted for a data time span ranging from 2011 to 2021, primarily due to the absence of certain data before 2011. Furthermore, despite the first BC pilot cities being included in 2014, it is imperative to incorporate data from preceding years to investigate the impacts of the policy, especially for the parallel trend test. Therefore, the data in this study covers the period from 2011 to 2021 and includes samples from 245 prefecture-level cities in China. All data are from the China City Statistical Yearbook, the China Construction Statistics Yearbook, the website of the Ministry of Industry and Information Technology of China, and the Institute of Digital Finance at Peking University. To ensure the completeness of the data, a few missing data points were supplemented using the mean value method. Table 1 shows the descriptive statistics.

Table 1 Descriptive statistics of the main variables.
Full size table

Empirical analysis

Parallel trend test

Figure 2 presents the result of the parallel trend test. The “current” in this diagram represents the year 2014, when the “Broadband China” demonstration cities were established. Meanwhile, “pre_n” and “post_n”, respectively, denote the years before and after the establishment year. Prior to the announcement of the BC pilot cities, some municipalities may prepare in response to receiving advanced information. Given that the unveiling of the BC strategy (in 2013) preceded the release of the pilot cities list (in 2014), advanced information is likely to disturb experimental outcomes. Therefore, the year 2013 (pre_1) was deleted from the regression, as the announcement of the BC strategy could affect the development of the cities in advance. No significant differences were found in the Theil index before the policy between the two groups, indicating the parallel trend test was passed. But after 2014, significant differences were investigated, indicating a significant impact of BC policy on the urban-rural income gap.

Fig. 2
figure 2

Parallel trend test.

Full size image

Baseline regression

Table 2 presents the results of the baseline regression analysis. From columns (1) to (6), control variables are incrementally incorporated into the regression Eq. (1). As a result of the gradual inclusion of control variables, it facilitates the detection of potential outliers. Additionally, this approach assists in evaluating the robustness of the baseline regression model. Meanwhile, the coefficients for BC remain positive and significant, indicating that BC policy leads to an increase in the Theil index. In other words, the Broadband China policy widens the income gap. This supports the hypothesis H1. Furthermore, the findings are consistent with the analysis of previous research and theoretical hypotheses (Nie and Wan, 2023; Deng et al., 2023). While the BC policy has led to an overall increase in residents’ income, urban residents have experienced greater income growth compared to rural residents (Nie and Wan, 2023). In terms of control variables, the urbanization rate (UTR), the level of foreign investment (FDI), the real GDP (RGDP), and the level of transportation facilities (ROAD) have contributed to the reduction of the urban-rural income disparity. Hence, it can be deduced that the BC policy is likely to yield greater economic benefits in rural areas in the foreseeable future. With the growth of GDP, urbanization, infrastructure, and so on, it is anticipated that the disparity in income between urban and rural areas will continue to diminish.

Table 2 Baseline regression analysis.
Full size table

Robustness test

Although some characteristic variables of the cities have been controlled, there is still a possibility that unobserved factors may affect the regression results. Constructing a “pseudo-BC variable” can serve to examine whether factors other than policies contribute to the widening income gap. In other words, if changes in the income gap are not influenced by the broadband China policy, the results of this pseudo-policy variable should align closely with the original baseline regression (Ding et al., 2024). Therefore, to verify whether the impact is incidental or not, this study conducts the placebo test by randomly sampling the data 500 times to construct the pseudo-treatment group. As shown in Fig. 3, the random regression results tend to have a normal distribution, with a small portion being significant. Meanwhile, the coefficient of the new BC is predominantly centered around 0 in the 500 regressions. As a result, it indicates that the impact of the BC policy is not a coincidence. The widening of the urban-rural income gap is indeed attributed to BC policy influences.

Fig. 3
figure 3

Placebo test.

Full size image

Table 3 presents the results of additional robustness tests. To minimize potential interference from the implementation of other relevant policies, this study controls the Big Data Comprehensive Pilot Zones policy (BDCPZ), the National Smart City pilot policy (SMART), and the Pilot policy of Information benefiting people (BENEFITING). These policies leverage digitalization to drive economic development, enhance public service delivery, and optimize resource allocation. The results are presented in column (1). After controlling these policies, the BC policy continued to exert a substantial influence on the urban-rural income gap, with minimal fluctuations in the estimated coefficient.

Table 3 Robustness tests.
Full size table

Besides, the income gap can also be represented by the ratio of urban income to rural income in the literature (Wang et al., 2024; Xia et al., 2024). Hence, to ensure the robustness of the results, the Theil index is substituted with the ratio of urban income and rural income (GAP2), and the regression result in column (2) indicates that the initial results were solid. Also, provincial capitals and municipalities normally possess greater economic power, a more robust political culture, and exert influence over surrounding cities. Moreover, these cities have a higher probability of being selected as pilot cities. This preference potentially impacts the initial findings. Therefore, this study conducts the analysis by using samples without provincial capitals and municipalities, and the results are in column (3) in Table 3. The coefficient for BC only experienced a slight decrease, which supports the initial findings.

Last, the selection of demonstration cities is not purely random. Some specific economic and social factors, such as GDP, income, consumption, culture, geographical location, and so on, may contribute to the selection of these cities as pilot cities. These non-random factors could potentially generate bias in the results. Hence, to address the endogenous problems due to selection bias, this study employs the nearest neighbor matching PSM-DID model and the results are in column (4) of Table 3 (Nepal et al., 2024). The results in column (4) show only a slight decrease in the coefficient of BC, providing strengthened support for hypothesis H2. All these results confirm the robust widening of the urban-rural income gap by the BC policy.

Regional heterogeneity

The BC strategy explicitly outlines different priority tasks based on regional distribution. The eastern regions have a higher demand for network upgrading and innovation, while the central and western regions are inclined towards broadband network development. As a result, the eastern provinces, central and western provinces are categorized into two heterogeneous groups based on the government’s division. As shown in columns (1) and (2) of Table 4, the impact of income gap expansion is more pronounced in the eastern provinces.

Table 4 Regional heterogeneity analysis.
Full size table

Furthermore, the BC policy emphasizes broadband network development in accordance with local circumstances. Compared to other cities, the Yangtze River Delta City Group, the Pearl River Triangle City Group, and the Bohai Rim Region are more developed. China arranges these three city clusters from south to north, each possessing unique characteristics. Hence, utilize them as exemplars for heterogeneous analysis. As indicated in columns (3) to (5) of Table 4, the income gap in the Pearl River Triangle City Group has widened significantly, while the effect of widening in the Yangtze River Delta City Group is the least pronounced.

Mechanism testing

This section conducts the mechanism tests according to Eqs. (2) and (3). The Broadband China policy affects the urban-rural income gap by influencing innovation, entrepreneurship, digital inclusive finance, and information sectors. Referring to Yu et al. (2021) and Zhang et al. (2024), this study employs the gradual regression model to assess the mediating effect.

Innovation and entrepreneurship

In Table 5, column (1) presents the analysis result with Eq. (1), which aligns with column (7) of Table 2. The results in column (2) indicate that the BC policy promotes the innovation level (INN), consistent with previous findings of Liu et al. (2023). Furthermore, column (3) demonstrates that the coefficients of INN and BC are positive, suggesting that BC policy exacerbates the income gap by elevating levels of innovation. Importantly, the coefficient of BC in column (3) is smaller than that in column (1), indicating a partial mediating effect. Besides, the analysis uses the entrepreneurial level as a mediator variable, with the results presented in columns (4) and (5). The results in column (4) indicate that the BC policy has significantly promoted entrepreneurship (ENT) in demonstration cities, consistent with theoretical assumptions. Similarly, the results in column (5) show that the BC policy has contributed to widening the income gap by promoting entrepreneurship, serving as a partial mediating effect. In conclusion, all the results in Table 5 provide support for hypothesis H2.

Table 5 Mechanism testing (Innovation and entrepreneurship).
Full size table

Digital-related Industries

In Table 6, column (1) is the same as column (7) of Table 2. As indicated in column (2), the BC policy has a positive impact on digital inclusive finance (DFI). Besides, the coefficients of DIF and BC in column (3) are positive, indicating that the BC policy exacerbates the income gap by increasing digital inclusive finance. Importantly, the coefficient of BC in column (3) is smaller than that in column (1) and significant at the 1% level, suggesting a partial mediating effect. Besides, the results in column (4) demonstrate that the BC policy has significantly contributed to the development of the information industry (PTS) in demonstration cities at a 1% level. Furthermore, column (4) indicates that the BC policy has widened the income gap by boosting the information industry. In conclusion, hypothesis H3 is established. BC policy has influenced the income gap through its promotion of digital-related industries.

Table 6 Mechanism testing (Digital-related industries).
Full size table

Policy synergies

To examine the potential impact of additional policies on the efficacy of BC policy, this study formulates Eq. (4). The interaction terms (BC_ECOMMERCE, BC_TECHFIANCE, BC_CONSUMPTION) serve as an indicator of whether a city is designated as a pilot city for two policies. In other words, the interaction terms are the dummy variables. When a city adopts BC policy in conjunction with another policy (ECOMMERCE, TECHFIANCE, or CONSUMPTION), the dummy variable equals 1, otherwise 0. Table 7 displays the regression results of Eq. (4). It indicates that the interaction terms of BC policy and the other three policies are all significantly positive in columns (1) to (3). Simultaneously, the coefficients of each interaction term are smaller than the coefficient of BC. Consequently, these three policies have effectively mitigated the widening impact of broadband China policy on the urban-rural income gap (Liu et al., 2024). This suggests that the “national e-commerce demonstration cities”, “national information consumption pilot cities”, and “the pilot policy of combining science and technology with finance” facilitate the equitable distribution of broadband construction among all residents. As a result, hypothesis H4 is supported. This clearly demonstrates the need for policy and industrial support for the economic value of broadband network construction.

Table 7 Policy synergies tests.
Full size table

Conclusions, implications, and recommendations

Conclusions

Income inequality has long been a pivotal concern in the realms of economic and social progress. In the era of information, there has been an increasing focus on examining the correlation between digitalization and income disparity (Qiu et al., 2021). The Broadband China policy has a significant impact on the urban-rural income gap, according to the literature, but there is a lack of in-depth studies, particularly in terms of mechanisms and policy synergies (Deng et al., 2023; Nie and Wan, 2023). Therefore, this study contributes to the impact of BC policy on the urban-rural income gap by incorporating additional evidence. More significantly, this study undertakes an investigation into the mechanisms and takes into account the influence of additional policies. This study provides valuable insights for scholars and policymakers to further investigate the policy effects. During the process of broadband construction, it is essential to focus on innovation, finance, industries, and relevant aspects while integrating them with appropriate policies to optimize the economic impact of broadband deployment.

Utilizing panel data from 245 cities from 2011 to 2021, this study investigates the impact of Broadband China policy on the urban-rural income gap and the mechanisms by employing the DID with multiple periods. The findings indicated that the BC policy has widened the income gap, which is robust under different tests. Furthermore, because of location-specific conditions, there is regional heterogeneity. Meanwhile, innovation, entrepreneurship, digital inclusive finance, and information industry development are essential mechanisms through which the BC policy influences the income gap. Importantly, strategies like National E-commerce Demonstration Cities, National Information Consumption Pilot Cities, and Pilot Zones of Combining Science and Technology with Finance should be considered. Due to the policy synergies, they enhance the advantages of the Broadband China policy. The synergistic effects between policies have weakened the trend of widening the urban-rural income gap. In summary, this study not only enhances the understanding of the relationship between BC policy and the urban-rural income gap but also contributes to research on policy synergy.

Implications

On the theoretical plane, the research outcomes of this study demonstrate that the policies associated with the digital economy exert an impact on the income distribution between urban and rural areas. Simultaneously, this aligns with certain literature, which validates that the development of the digital economy ought to take into account the coordinated advancement of urban and rural areas (Gao and Li, 2023; Peng and Dan, 2023). Through the mechanism test, this study discovers more pathways by which the BC policy influences the urban-rural income gap, such as innovation, entrepreneurship, and digital finance. This fills a portion of the void in the study of the digital economy and urban-rural development. More significantly, the study should not be confined to a single policy but rather systematically deliberate on complex factors. This study fills the theoretical gap regarding the synergy effect of the Broadband China (BC) policy in the existing literature and combines the BC policy with other national strategies. The results also demonstrate that the combination of multiple policies is conducive to the exertion of the BC policy effect.

At the practical level, the outcomes of this study thoroughly demonstrate that the implementation of the BC policy ought to be based on actual local circumstances and entail systematic deliberations to facilitate the common development of urban and rural areas. Simultaneously, it is beneficial for policymakers to take the disparities between urban and rural areas into more comprehensive account. Besides, the empirical results of this study demonstrate that the cities are in need of introducing certain industrial supports for digitization. Particularly in domains, it can facilitate the BC policy to exert more desirable policy effects, such as in innovation and finance. More significantly, the findings regarding the policy synergy effect fully assist policymakers in realizing the significance of the coordination of multiple policies.

Recommendations

This study provides the following policy recommendations: First, adequately execute the relevant mandates of the BC policy, with a particular emphasis on maximizing the economic value of broadband networks. The baseline regression shows the widening of the urban-rural income gap because of the BC policy. Referring to the existing literature, it may be due to the income growth of urban residents being more significant than that of rural residents (Nie and Wan, 2023; Liu et al., 2024). Therefore, the impact of BC policy on residents’ income and economic development deserves recognition. In the meantime, the urbanization rate, the foreign investment, the real GDP, and the level of transportation facilities are conducive to reducing the income gap. Hence, it is imperative to consider such factors during the implementation of broadband infrastructure. By promoting equalization of services, it stimulates rural economic development and reduces the income gap.

Second, the implementation of construction systems suitable for local development is crucial. The results of regional heterogeneity revealed varying policy effects in BC pilot cities across regions. When promoting construction demonstrations, relevant departments should consider the levels of local broadband construction, service application, and development environment. Therefore, as a guide for future development, it is essential to build an evaluation index system for the level of broadband network construction. Meanwhile, the governments should also assume a comprehensive role in the allocation of market resources.

Third, encourage regional innovation and industrial integration, particularly in support of rural entrepreneurship. The results of the mechanism testing fully illustrated that broadband networks create economic value by stimulating innovation, entrepreneurship, and related industrial development. Therefore, when implementing BC policy, it is imperative to prioritize the promotion of local innovation and entrepreneurship. At the same time, financial institutions should offer tailored digital financial services and integrate policies to drive advancements in science and technology finance. Simultaneously, adolescents should be encouraged to engage in employment or entrepreneurship in their hometowns. This will facilitate the advancement of innovative technologies, the integration of urban-rural infrastructure networking, and the improvement of industrial service capabilities.

Fourth, local government should consider the synergistic impact of implementing multiple policies to create an enhanced policy environment. In this study, the synergistic effects between policies have significantly weakened the widening trend of the income gap. If the construction of broadband networks would like to have a significant economic impact, they must be supported by appropriate policies. Hence, in the process of constructing Broadband China demonstration cities, it is important for the government to consider the synergistic impact of different strategies. More importantly, the implementation of different policies should align with practicality and serve the best interests of the populace.

This study also has some limitations that need to be considered in further studies: First, the multi-period DID method may not be the best for analyzing the phased effects of digital infrastructure on the income gap. Second, due to data availability constraints, the analysis of potential research content may be incomplete. Third, due to spatial constraints, this paper does not provide an exhaustive categorization of industries and regions. It is hoped that this study will inspire other researchers with new ideas for relevant studies.