Does New-Type Consumption Enhance Urban Economic Resilience? Evidence from China’s Information Consumption Pilot Policy

Abstract

Against the backdrop of frequent internal and external shocks, as a core driver of the consumption segment in the digital economy, the impact mechanism and actual effectiveness of information consumption on urban economic resilience urgently require systematic exploration. Based on panel data of 280 prefecture-level cities in China from 2010 to 2022, this study treats the information consumption pilot policy as a quasi-natural experiment and employs a multi-period Difference-in-Differences (DID) method to empirically examine the policy’s impact on urban economic resilience and its internal mechanisms. The results show that the information consumption pilot policy significantly enhances urban economic resilience, with a policy effect coefficient of 0.084, and this conclusion remains robust after multiple robustness tests. Mechanistic analysis indicates that the policy indirectly strengthens urban economic resilience by promoting consumption growth, stimulating technological innovation, and improving human capital. Meanwhile, the level of digital infrastructure plays a positive moderating role in the policy effect. Heterogeneity analysis finds that the policy has a more pronounced effect of enhancing economic resilience on cities with larger population sizes, higher economic density, and non-resource-dependent characteristics. Further extended research confirms that the information consumption pilot policy exhibits a significant spatial spillover effect on urban economic resilience, and this spillover effect presents a phased characteristic of “resource homogeneous competition → positive synergistic driving → cross-regional resource siphoning → spatial attenuation of the effect” with changes in geographical distance.

1. Introduction

The in-depth integration of globalization and the digital technology revolution is driving the world economy into a new phase of normalized shocks. The aftershocks of the 2008 international financial crisis have not yet subsided, while new shocks such as the COVID-19 pandemic and geopolitical conflicts have followed, leaving the world economy facing a complex situation of overlapping multiple pressures. Data from the International Monetary Fund (IMF) shows that the global economic growth rate plummeted from 4.9% to 3.2% in 2008, leading to a significant slowdown in the growth of regions dominated by traditional manufacturing [1]. In 2020, the COVID-19 pandemic dealt a heavy blow to service-oriented economies: China’s catering revenue and domestic tourism revenue dropped by 16.6% and 61.1% year-on-year, respectively [2,3]. These facts highlight the inherent vulnerability of economic systems relying on traditional development paths in terms of shock resistance. Against this background, exploring ways to enhance urban economic resilience is not only a passive defense against uncertain challenges but also a strategic choice to proactively reshape competitive advantages and achieve sustainable development [4].

Among the various paths to enhance economic resilience, the rise of information consumption provides a new approach to addressing the vulnerability of traditional development models. As a core consumption form in the digital economy era, information consumption refers to economic activities in which residents or enterprises purchase and use information products or enjoy information services to meet production and living needs. Its core lies in utilizing products and services related to information and communication technology (ICT) to satisfy various demands [5]. This consumption model not only directly drives economic growth but also holds great potential for enhancing urban economic resilience due to its close connection with technological innovation and industrial upgrading [6]. Since China launched information consumption pilot programs in 2013, the information consumption market has experienced explosive growth under policy guidance and market cultivation. Data shows that the market scale jumped from 1.8 trillion yuan to 6.8 trillion yuan between 2012 and 2021, with an average annual growth rate of over 15%; in the first half of 2024, the market scale exceeded 3 trillion yuan, a year-on-year increase of 6%, showing a strong development momentum. Regional practices also confirm its value: for example, online retail sales accounted for nearly 40% of total social retail sales of consumer goods in Beijing in the first three quarters of 2024, and the added value of the core digital economy industry in Zhejiang Province accounted for 12.3% of GDP, maintaining double-digit growth for 10 consecutive years. These facts not only reflect the direct driving effect of information consumption on the regional economy but also raise a deeper question: beyond promoting growth, can information consumption become a key support for enhancing urban economic resilience?

Although existing studies have recognized the enabling role of the digital economy in economic resilience [7,8], there is an obvious perspective bias that makes it difficult to fully address the above core question. On the one hand, most existing literature focuses on resilience mechanisms on the production side, such as the role of digital infrastructure [9], digital platform penetration [10], and industrial digital transformation [11,12], while relatively neglecting the important contribution of the consumption side. On the other hand, research on the information consumption pilot policy itself is mostly limited to its short-term economic performance effects, such as driving industrial upgrading [13], improving regional innovation efficiency [14], and promoting employment [15], while lacking systematic exploration of its potential value in enhancing urban economic resilience. This limitation in research perspective hinders a comprehensive understanding of the new mechanisms and paths for urban economic resilience in the digital economy era.

In view of this, this study uses panel data of 280 prefecture-level cities in China from 2010 to 2022, treats the information consumption pilot policy as a quasi-natural experiment, and adopts a multi-period DID method to systematically evaluate the impact of information consumption development on urban economic resilience and its mechanism of action. The marginal contributions of this study are mainly reflected in the following three aspects: First, it expands the research perspective. Breaking away from the focus on the production side in existing literature, this study starts from the consumption side and systematically examines the resilience-enhancing effect of the information consumption pilot policy on urban economic resilience, providing a new dimension for understanding how the digital economy empowers economic resilience. Second, it deepens the dimension of policy evaluation. This study expands the evaluation of the information consumption policy from short-term economic performance to the long-term structural indicator of urban economic resilience, enriching the theoretical connotation and practical significance of policy effect evaluation. Third, it clarifies the mechanism of action and boundary conditions. This study not only identifies the transmission paths of consumption growth, technological innovation, and human capital improvement but also verifies the moderating role of digital infrastructure. Through heterogeneity analysis and spatial spillover effect testing, it reveals the differentiated characteristics and spatial laws of policy effects, providing empirical evidence for targeted policy-making.

2. Policy Background and Research Hypotheses

2.1. Policy Background

In the wave of global digital economy development, information consumption, as a new driver for domestic demand growth, has increasingly prominent strategic status. To break through bottlenecks in infrastructure, innovation, and institutions during its development, the Chinese government has implemented a series of systematic policy interventions since 2013, and the policy evolution has exhibited distinct phased characteristics: The initial stage (2013–2015) focused on pilot exploration. By establishing national pilot cities for information consumption, the first batch of pilot areas was announced at the end of 2013, and the second batch at the end of 2014 (see

Table 1. List of Pilot Areas.

List of First-Batch (End of 2013) Pilot Cities for Information Consumption

List of Second-Batch (End of 2014) Pilot Cities for Information Consumption

Municipalities directly under the Central Government: Beijing, Tianjin, Chongqing, Changning District of Shanghai, Yangpu District of Shanghai Hebei Province: Shijiazhuang, Qinhuangdao, Tangshan, Yongnian County of Handan Shanxi Province: Taiyuan Liaoning Province: Dalian, Shenyang Jilin Province: Jilin City, Yanbian Korean Autonomous Prefecture, Changchun Jingyue High-tech Industrial Development Zone Heilongjiang Province: Harbin, Daqing Jiangsu Province: Nanjing, Yancheng, Zhangjiagang, Guangling District of Yangzhou Zhejiang Province: Ningbo, Hangzhou, Jinhua (Yiwu), Jiaxing Anhui Province: Hefei, Wuhu, Ma’anshan Fujian Province: Xiamen, Fuzhou, Shishi Jiangxi Province: Nanchang, Zhanggong District of Ganzhou Shandong Province: Weihai, Zibo, Jining, Weifang Henan Province: Zhengzhou, Jiyuan Guangdong Province: Shenzhen, Shantou, Zhuhai, Huizhou Hunan Province: Zhuzhou, Hengyang, Chenzhou Hubei Province: Wuhan, Xiangyang, Xiaonan District of Xiaogan Hainan Province: Haikou Sichuan Province: Chengdu, Mianyang, Nanchong, Leshan Guizhou Province: Xixiu District of Anshun, Honghuagang District of Zunyi Yunnan Province: Yuxi Shaanxi Province: Baoji Gansu Province: Lanzhou, Jiayuguan Qinghai Province: Xining, Golmud Inner Mongolia Autonomous Region: (None) Guangxi Zhuang Autonomous Region: Nanning, Liuzhou, Guilin Ningxia Hui Autonomous Region: Yinchuan Xinjiang Uygur Autonomous Region: Karamay, Yining

Municipality directly under the Central Government: Shanghai Hebei Province: Baigou New Town Shanxi Province: Changzhi Liaoning Province: Benxi Jilin Province: Hunchun, Baicheng Heilongjiang Province: Mudanjiang Jiangsu Province: Xuzhou, Suzhou Zhejiang Province: Shaoxing Anhui Province: Anqing, Bengbu Fujian Province: Quanzhou Jiangxi Province: Wuyuan County of Shangrao, Xinyu Shandong Province: Wendeng District of Weihai, Rencheng District of Jining Henan Province: Luoyang, Xinxiang Guangdong Province: Foshan Hunan Province: Wuling District of Changde Hubei Province: Huangshi Sichuan Province: Meishan Guizhou Province: Guiyang Yunnan Province: Dali, Baoshan Shaanxi Province: Xianyang Gansu Province: Baiyin, Dunhuang Qinghai Province: Delingha Inner Mongolia Autonomous Region: Ordos, Manzhouli Guangxi Zhuang Autonomous Region: Beihai Ningxia Hui Autonomous Region: Wuzhong, Yuanzhou District of Guyuan Xinjiang Uygur Autonomous Region: Korla

for detailed lists), so as to explore the development path in a “point-to-area” manner; The mid-term (2016–2020) shifted to improving quality and efficiency, with the policy focus moving from scale expansion to mode refinement and supply-side structural reform, thus constructing a more refined policy framework; The recent period (since 2021) has elevated information consumption to a national strategy, fully integrating it into the top-level design such as the 14th Five-Year Plan, with policy objectives focusing on ecosystem construction and consumption scenario innovation. This gradual policy evolution from “pilot exploration” to “strategic integration” not only provides an institutional guarantee for the sustained growth of information consumption but also forms a quasi-natural experiment with rich layers and a long time span. The pilot cities in different batches and stages differ in policy intervention intensity and content, which provides an ideal empirical scenario and a solid institutional foundation for this study to accurately identify the net effect and its mechanism of information consumption policies on urban economic resilience by using quantitative methods such as the multi-period difference-in-differences (DID) method.

2.2. Research Hypotheses

2.2.1. Direct Impact of the Information Consumption Pilot Policy on Urban Economic Resilience

Urban economic resilience refers to the buffering, adaptation, and transformation capabilities demonstrated by the urban economic system when facing internal and external shocks. As an emerging consumption form, information consumption can directly act on and enhance urban economic resilience through its characteristics of demand rigidity, industrial relevance, and technological intensity.

(1)

Demand rigidity: Strengthening buffering capacity

With the in-depth advancement of digital transformation in modern economic society, digital technology has been deeply integrated into the core scenarios of production and daily life [16]. On the production side, the full-process optimization of enterprise digital operations increasingly relies on the technical support of industrial software and the computing power empowerment of cloud services, and the path dependence of production organization models and resource allocation efficiency on digital technology continues to strengthen [17]; on the daily life side, scenario innovations such as the breakthrough of spatial boundaries in residents’ daily social interactions and the improved accessibility of smart medical services are promoting the evolution of individuals’ functional dependence on the digital ecosystem toward in-depth integration. The digital scenario has become a key infrastructure for maintaining social operation and people’s livelihood security [18].

Such digital demands usually exhibit counter-cyclical fluctuation characteristics, and their scale is not prone to significant contraction due to short-term economic fluctuations or external shocks. During an economic downturn, traditional consumption may contract due to adjustments in income expectations, but enterprises’ demand for investment in digital operation guarantees such as data storage and network security, as well as residents’ consumption demand for digital life services such as online education and telecommuting, can often remain relatively stable. This demand rigidity of information consumption provides a stable underlying support for urban economic resilience. At the same time, information consumption pilot cities further release the potential of such rigid demands by systematically improving information infrastructure and continuously optimizing the consumption environment [19], making it a stabilizer for economic growth. When external shocks weaken the growth momentum of traditional industries, the stable demand for information consumption can directly support the total economic scale, reduce the magnitude of macroeconomic fluctuations, and provide important buffer space for the urban economy to resist external shocks, thereby strengthening its systematic buffering capacity.

(2)

Industrial relevance: Enhancing adaptability

Information consumption is not an isolated consumption form but has the ecological linkage attribute of “demand-driven → supply response → chain expansion”, enabling it to build a complete industrial value cycle system [20]. Specifically, residents’ consumption demand for smart terminals can drive the scale expansion and technological iteration of hardware industries such as chip manufacturing and electronic assembly through market transmission; enterprises’ procurement demand for cloud services can drive technological upgrading and service innovation in software industries such as cloud computing and big data; digital content consumption can activate creative output and model innovation in related fields such as cultural and creative industries and media.

This extensive industrial network, supported by policy tools such as targeted industrial chain subsidies and innovation platform construction in information consumption pilot cities to accurately support upstream and downstream enterprises, is gradually promoting the transformation of the urban economic structure from single dependence on traditional industries to a diversified pattern led by information consumption and coordinated development of multiple industries [21]. When specific traditional industries face growth difficulties due to external shocks (such as increased resource constraints and shrinking market demand), the related industries driven by information consumption can fill the economic gap through a substitute growth effect. For example, during the period when offline retail scenarios were restricted, the scale expansion of information consumption-related industries such as live-streaming e-commerce and instant delivery effectively offset losses in the traditional consumption sector. This industrial relevance of information consumption significantly enhances the adaptive flexibility of the urban economy by driving the diversified restructuring of the economic structure. This dynamic flexibility formed based on structural adjustment is the core manifestation of the urban economic system’s ability to adapt to changes in the external environment.

(3)

Technological intensity: Promoting recovery and transformation capabilities

As an important carrier of technological innovation and application, the development of information consumption is essentially a process of continuous iteration of digital technology and continuous innovation of application scenarios. Under the guidance and incentive of the information consumption pilot policy, the commercialization and implementation of cutting-edge technologies such as 5G, artificial intelligence, and the Internet of Things in consumption scenarios have accelerated significantly. This not only gives rise to breakthrough emerging business formats such as intelligent driving and the metaverse but also penetrates deeply into traditional industries through technology diffusion mechanisms [22,23]. Specifically, manufacturing realizes intelligent restructuring of production processes relying on the industrial Internet, agriculture promotes the transformation of planting models to precision through big data technology, and the service industry achieves systematic innovation in service forms and delivery models through digital technology empowerment.

The digital assets and capability reserves accumulated through such technological applications are transformed into core driving forces for economic recovery after the occurrence of external shocks. For example, telecommuting technology quickly adapted to enterprise operation needs after the pandemic shock, and smart logistics systems efficiently restored the material allocation network after disaster events. This rapid response mechanism formed based on previous technological reserves significantly improves the efficiency of the urban economy in achieving recovery and transformation from shocks.

In summary, the demand rigidity of information consumption strengthens the buffering capacity of the urban economy, its industrial relevance enhances the adaptability, and its technological intensity promotes the recovery and transformation capabilities. Based on this, the following hypothesis is proposed:

Hypothesis H1.

The information consumption pilot policy can significantly enhance urban economic resilience.

2.2.2. Indirect Impact of the Information Consumption Pilot Policy on Urban Economic Resilience

(1)

Consumption growth effect

The information consumption pilot policy focuses on activating demand, and drives the consumption upgrading of residents and enterprises through subsidies, scenario cultivation, and product innovation [24]. On the resident side, subsidies for smart terminals and discounts on digital services reduce the consumption threshold, stimulating the growth of consumption of information products and services; on the enterprise side, policy incentives encourage enterprises to increase digital investment and expand the procurement of tools such as information systems and industrial software, which not only improves operational efficiency but also expands the demand for information consumption-related technologies and equipment.

When the urban consumption level improves, on the one hand, the expansion of consumption scale directly drives the development of upstream and downstream industries in the information consumption industrial chain—from hardware manufacturing to software services, from content production to platform operation—forming a multiplier effect for industrial development, promoting economic growth and industrial structure optimization, and enhancing the diversity and risk resistance of the urban economy. On the other hand, high-level information consumption gives rise to new consumption demands and business models, such as personalized consumption recommendations based on big data analysis and smart life services relying on the Internet of Things, driving innovation in the consumer market. This enables the urban economy to have stronger adaptability and adjustability when facing changes in consumption trends and external shocks, thereby enhancing urban economic resilience [25].

(2)

Human capital improvement effect

The implementation of the information consumption pilot policy improves the urban human capital level from multiple dimensions [26]. First, the policy drives the development of the information consumption industry, creating a large number of high-skill jobs, attracting the inflow of high-quality external talents, and at the same time promoting the transfer of local talents to the information consumption field for employment. Second, to meet the needs of industrial development, the policy promotes information consumption pilot cities to strengthen the construction of vocational education and continuing education systems, optimize curriculum settings and improve teaching quality for information consumption-related majors, and cultivate locally adapted talents; enterprises also increase internal training investment to improve employees’ digital skills and innovation capabilities.

When the urban human capital level improves, on the one hand, high-quality talents provide intellectual support for the development of the information consumption industry, driving industrial innovation and upgrading and enhancing industrial competitiveness [27]. On the other hand, the spillover effect of human capital radiates to other industries in the city, improving overall labor productivity and promoting collaborative innovation and integrated development among industries [28]. At the same time, cities with high human capital levels have more advantages in attracting innovative resources and undertaking the layout of emerging industries, and can cultivate more economic growth points. When facing economic shocks, relying on talent advantages and diversified industrial support, the urban economy can more easily achieve recovery and transformation, thereby enhancing economic resilience [29].

(3)

Technological innovation effect

Through targeted policy tools such as establishing special R&D subsidies, building industry-university-research collaborative innovation platforms, and improving the full-chain intellectual property protection system, the information consumption pilot policy effectively encourages market entities and research institutions to increase investment intensity in technological R&D in the information field. Empowered by policies, enterprises in pilot cities focus on targeted research in key areas such as 5G technology scenario-based application innovation, artificial intelligence core algorithm optimization, and big data key processing technology breakthroughs; universities and research institutes also deepen basic research and achievement transformation applications related to information consumption, forming a closed loop of technological R&D and achievement implementation, thereby systematically improving the urban technological innovation capability [30].

The continuous improvement of urban technological innovation capability will lay a solid resilience foundation for resisting external shocks [31]. On the one hand, innovative achievements are directly transformed into new products and service forms in the information consumption field, such as intelligent terminal equipment, high-efficiency information processing systems, and immersive digital content, continuously optimizing the supply quality and market attractiveness of information consumption, and strengthening the supporting role of the consumption side in the economy [32]. On the other hand, technological innovation in the information field radiates and diffuses to other industries in the city through industrial chain penetration and technology spillover effects, accelerating the digital transformation process of traditional industries, and significantly improving the production efficiency and core competitiveness of various industries, consolidating the risk resistance of the industrial side [33]. When facing external pressures such as technological blockades and market fluctuations, relying on solid advanced technology reserves and a diversified industrial structure, the urban economy can more quickly realize resource reorganization, structural adjustment, and functional recovery, thereby systematically enhancing economic resilience from the perspective of technological support [34].

In summary, the following hypothesis is proposed:

Hypothesis H2.

The information consumption pilot policy can indirectly enhance urban economic resilience through the consumption growth effect, human capital improvement effect, and technological innovation effect.

2.2.3. Moderating Role of Digital Infrastructure

As a key link connecting information consumption and urban economic resilience, the development level of digital infrastructure significantly moderates the resilience-enhancing effect of the information consumption policy on the economy.

First, information consumption relies on digital infrastructure [35]. A low level of digital infrastructure will form technical bottlenecks, restricting the popularization and deepening of information consumption, and weakening its direct and indirect role in promoting economic resilience. Conversely, improved infrastructure can promote the “technology → industry → consumption” cycle of information consumption, more effectively stimulating the market and driving upgrading, thereby strengthening its role in enhancing economic resilience [36]. Second, digital infrastructure has network externalities [37]. When the infrastructure level is low, network nodes are sparse, and the synergy effect and overall improvement effect of information consumption are limited. When the infrastructure reaches a certain scale, the network effect emerges: information flow accelerates, innovation diffusion expands, which not only enhances the risk resistance of related industries but also improves the resilience of traditional industries through technology spillovers, significantly amplifying the positive impact of information consumption on economic resilience [38]. Finally, improved infrastructure can reduce the costs of information transmission and market transactions. This reduces frictions in information consumption (such as payment security and data sharing barriers) and increases market participation. At the same time, low-cost information flow accelerates the optimal allocation of factors, making it easier for the growth potential of information consumption to be converted into economic resilience. Conversely, weak infrastructure increases transaction costs, inhibits factor flow, and weakens transmission efficiency [39].

Based on this, the following hypothesis is proposed:

Hypothesis H3.

Digital infrastructure plays a positive moderating role in the impact of information consumption on urban economic resilience.

In summary, the theoretical framework of this study is shown in Figure 1:

3. Research Design

3.1. Model Specification

(1)

Baseline Regression Model

To identify the causal effect of the information consumption pilot policy on urban economic resilience, this study treats the information consumption pilot policy as a quasi-natural experiment and constructs a progressive DID model for empirical research. The model is constructed as follows:

$${\mathrm{Er}}_{\mathrm{it}}={\mathrm{\alpha }}_0+{\mathrm{\alpha }}_1\mathrm{Policy}+{\mathrm{\alpha }}_2{\mathrm{Controls}}_{\mathrm{it}}+{\mathrm{\mu }}_{\mathrm{i}}+{\mathrm{\lambda }}_{\mathrm{t}}+{\mathrm{\epsilon }}_{\mathrm{it}}$$

(1)

In the formula, ${\mathrm{Er}}_{\mathrm{it}}$ represents the economic resilience level of city $\mathrm{i}$ in year $\mathrm{t}$; $\mathrm{Policy}={\mathrm{treat}}_{\mathrm{i}}\times {\mathrm{post}}_{\mathrm{t}}$ is the core explanatory variable of this study: if the city where i is located is one of the first or second batch of information consumption pilot cities approved by the Ministry of Industry and Information Technology, the policy dummy variable ${\mathrm{treat}}_{\mathrm{i}}$ takes a value of 1, otherwise 0; Given that the two batches of pilot areas were announced at the end of 2013 and the end of 2014, respectively, the dummy variable ${\mathrm{post}}_{\mathrm{t}}$ is defined as follows: For the first batch of pilot cities, it takes a value of 0 before 2014 and 1 in 2014 and thereafter; For the second batch of pilot cities, it takes a value of 0 before 2015 and 1 in 2015 and thereafter; represents ${\mathrm{Controls}}_{\mathrm{it}}$ variables; ${\mathrm{\mu }}_{\mathrm{i}}$ and ${\mathrm{\lambda }}_{\mathrm{t}}$ represent individual and time fixed effects, respectively; ${\mathrm{\alpha }}_0$ is the intercept term; ${\mathrm{\alpha }}_1$ is the regression coefficient of the explanatory variable, which reflects the impact of the information consumption pilot policy on urban economic resilience; ${\mathrm{\alpha }}_2$ is the regression coefficient of the control variables; ${\mathrm{\epsilon }}_{\mathrm{it}}$ is the random error term.

(2)

Mediation Effect Model

To verify Hypothesis H2, this study refers to the method of Jiang Ting [40] and constructs the following mediation effect model:

$${\mathrm{M}}_{\mathrm{it}}={\mathrm{\beta }}_0+{\mathrm{\beta }}_1\mathrm{Policy}+{\mathrm{\beta }}_2{\mathrm{Controls}}_{\mathrm{it}}+{\mathrm{\mu }}_{\mathrm{i}}+{\mathrm{\lambda }}_{\mathrm{t}}+{\mathrm{\epsilon }}_{\mathrm{it}}$$

(2)

In the formula, ${\mathrm{M}}_{\mathrm{it}}$ is the mediating variable (including consumption growth, technological innovation, and human capital); ${\mathrm{\beta }}_0$ are constants; ${\mathrm{\beta }}_1$, ${\mathrm{\beta }}_2$ are coefficients; other variables are the same as in Formula (1).

(3)

Moderating Effect Model

To verify Hypothesis H3, this study constructs the following moderating effect model:

$${\mathrm{Er}}_{\mathrm{it}}={\mathrm{\gamma }}_0+{\mathrm{\gamma }}_1\mathrm{Policy}\times {\mathrm{Di}}_{\mathrm{it}}+{\mathrm{\gamma }}_2\mathrm{Di}+{\mathrm{\gamma }}_3{\mathrm{Controls}}_{\mathrm{it}}+{\mathrm{\mu }}_{\mathrm{i}}+{\mathrm{\lambda }}_{\mathrm{t}}+{\mathrm{\epsilon }}_{\mathrm{it}}$$

(3)

In the formula, $\mathrm{Di}$ is the moderating variable (digital infrastructure construction level); ${\mathrm{\gamma }}_0$ is a constant; ${\mathrm{\gamma }}_1$, ${\mathrm{\gamma }}_2$, ${\mathrm{\gamma }}_3$ are coefficients; other variables are the same as in Formula (1).

(4)

Spatial Durbin Model

To further verify the spatial spillover effect of the information consumption pilot policy on urban economic resilience, this study constructs the following spatial econometric model:

$$\begin{array}{c}\mathrm{E}{\mathrm{R}}_{\mathrm{it}}={\mathrm{\gamma }}_0+{\mathrm{\gamma }}_1\mathrm{Policy}+{\mathrm{\gamma }}_2\mathrm{W}\times \mathrm{Policy}+\mathrm{\rho }\mathrm{W}\times \mathrm{E}{\mathrm{R}}_{\mathrm{it}}+{\mathrm{\gamma }}_3{\mathrm{Controls}}_{\mathrm{it}}+\\ {\mathrm{\gamma }}_4\mathrm{W}\times {\mathrm{Controls}}_{\mathrm{it}}+{\mathrm{\mu }}_{\mathrm{i}}+{\mathrm{\lambda }}_{\mathrm{t}}+{\mathrm{\epsilon }}_{\mathrm{it}}\end{array}$$

(4)

In the formula, ${\mathrm{\gamma }}_0$ is a constant; ${\mathrm{\gamma }}_1$, ${\mathrm{\gamma }}_2$, ${\mathrm{\gamma }}_3$, ${\mathrm{\gamma }}_4$ are coefficients; $\mathrm{W}$ is the spatial weight matrix, and adjacency, inverse distance, and economic distance weight matrices are used, respectively, to reflect the impact of the information pilot policy on the economic resilience of neighboring regions; $\mathrm{W}\times \mathrm{Policy}$, $\mathrm{W}\times {\mathrm{ER}}_{\mathrm{it}}$, and $\mathrm{W}\times {\mathrm{Controls}}_{\mathrm{it}}$ are the spatial lag terms of the pilot policy, urban economic resilience, and control variables, respectively; $\mathrm{\rho }$ is the coefficient of the spatial lag term; other variables are the same as in Formula (1).

3.2. Variable Definition

(1)

Dependent Variable: Urban Economic Resilience (Er). Current academic research on the measurement of economic resilience mainly includes two categories: single-indicator method and indicator system method. Among the single-indicator methods, the sensitivity index framework proposed by Martin [41] is widely recognized, which quantifies resilience by measuring the deviation between actual economic fluctuations and expected trends. The indicator system method constructs a comprehensive evaluation system covering multiple dimensions such as risk resistance, adaptation, and recovery and regeneration [42].

However, this study argues that the indicator system method has significant limitations: first, the subjectivity of indicator selection and weight setting leads to a lack of comparability between studies; second, if the system includes driving factors, it is easy to confuse causal relationships; third, there may be a large deviation between the measured value and the actual resilience level, affecting the accuracy of conclusions [43]. Based on this, this study uses the single-indicator method to measure urban economic resilience. The calculation method of economic resilience based on the national real GDP growth rate is shown in Formula (5). In addition, to enhance the robustness of the conclusions, the indicator system method will be used to measure urban economic resilience in the robustness test for supplementary verification.

$${\mathrm{ER}}_{\mathrm{it}}={\displaystyle \frac{{\mathrm{\Delta }\mathrm{Y}}_{\mathrm{it}}-{\mathrm{\Delta }\mathrm{E}}_{\mathrm{it}}}{\left|{\mathrm{\Delta }\mathrm{E}}_{\mathrm{it}}\right|}}={\displaystyle \frac{{\mathrm{Y}}_{\mathrm{i},\mathrm{t}-1}\times { \mathrm{g}}_{\mathrm{it}}-{\mathrm{Y}}_{\mathrm{i},\mathrm{t}-1}\times { \mathrm{e}}_{\mathrm{it}}}{\left|{\mathrm{Y}}_{\mathrm{i},\mathrm{t}-1}\times { \mathrm{e}}_{\mathrm{it}}\right|}}={\displaystyle \frac{{\mathrm{g}}_{\mathrm{it}}-{\mathrm{e}}_{\mathrm{it}}}{\left|{\mathrm{e}}_{\mathrm{it}}\right|}}$$

(5)

In the formula, ∆${\mathrm{Y}}_{\mathrm{it}}$ represents the actual GDP change of city i in year t; ∆${\mathrm{E}}_{\mathrm{it}}$ represents the expected GDP change of city $\mathrm{i}$ in year $\mathrm{t}$; ${\mathrm{Y}}_{\mathrm{i},\mathrm{t}-1}$ is the actual GDP of city $\mathrm{i}$ in year $\mathrm{t}$; ${\mathrm{g}}_{\mathrm{it}}$ is the actual GDP growth rate of city i in year $\mathrm{t}$; ${\mathrm{e}}_{\mathrm{it}}$ is the expected GDP growth rate of city i in year $\mathrm{t}$, measured by the national real GDP growth rate in year $\mathrm{t}$. A larger value indicates that city i has stronger economic resilience relative to the national average level.

(2)

Core Explanatory Variable: Information Consumption Pilot Policy ($\mathrm{Policy}$). Based on the list of two batches of information consumption pilot cities announced by the Ministry of Industry and Information Technology, this study constructs a pilot dummy variable (${\mathrm{treat}}_{\mathrm{t}}$) and a pilot time dummy variable (${\mathrm{post}}_{\mathrm{i}}$), and determines their interaction term ($\mathrm{Policy}={\mathrm{treat}}_{\mathrm{t}}\times { \mathrm{post}}_{\mathrm{i}}$) as the core explanatory variable. Considering that the actual implementation time of the first and second batches of information consumption pilot policies was the end of 2013 and the end of 2014, respectively, 2014 and 2015 are set as the time nodes when pilot regions began to be affected by the policy. In addition, to meet the assumption of individual treatment stability in the DID method and ensure the accuracy of regression analysis results, cities where the pilot program was only carried out in county-level administrative units or a certain district under a prefecture-level city are excluded from the sample. Finally, 50 pilot cities are determined as the treatment group, and 230 non-pilot cities as the control group.

(3)

Control Variables: To control the impact of other factors on urban economic resilience, referring to relevant studies [44], this study introduces the following control variables: Industrial Structure (Str): Measured by the ratio of the added value of the tertiary industry to the gross regional product; Opening-up Level (Ope): Measured by the proportion of total import and export goods to regional GDP; Economic Development Level (Ed): Measured by per capita regional GDP, with a logarithm taken; Urbanization Rate (Cl): Measured by the proportion of urban permanent population to the total permanent population; Fiscal Decentralization Degree (Fis): It is represented by the proportion of general government fiscal revenue to general government fiscal expenditure.

(4)

Mechanism Variables: Consumption Growth (Con): Referring to the study of Huang Qinghua and Xiang Jing [45], per capita total retail sales of social consumer goods is used to measure the growth of consumption quality and quantity; Human Capital (Pc): Referring to the method of Zhan Xinyu and Liu Wenbin [46], the number of students enrolled in ordinary institutions of higher learning in a prefecture-level city as a proportion of the total regional population is used to represent human capital; Technological Innovation (Tc): Referring to the study of Yin Tianbao et al. [47], the number of patent grants per 10,000 people is used to represent the regional technological innovation level; Digital Infrastructure Construction Level (Di): Measured with reference to the method proposed by Chao Xiaojing et al. [48], this indicator follows a three-step process: first, systematically collecting government work reports from 280 prefecture-level cities nationwide spanning 2010–2022, and delineating the scope of 51 digital infrastructure-related terms included in the statistics—covering domains such as 5G, mobile communication, and information technology; second, conducting text segmentation on these reports using Python 3.13.8, removing meaningless stop words, then separately counting the total word count of each report and the frequency of terms related to new digital infrastructure; finally, using the ratio of new digital infrastructure-related term frequency to the total word count of the report to quantitatively characterize the development level of new digital infrastructure in the corresponding prefecture-level city.

3.3. Data Description and Descriptive Statistics

Considering the availability and completeness of data, this study selects 280 prefecture-level cities across China from 2010 to 2022 as the research sample. The indicator data required for the empirical analysis are mainly obtained from authoritative release channels such as the China City Statistical Yearbook, local government statistical bulletins, the official database of the National Bureau of Statistics, and local government portals. For a small amount of missing data during the sample period, linear interpolation is used to fill in the gaps; to reduce the interference of outliers on the estimation results, all continuous variables are winsorized at the 1% and 99% quantile levels. After data preprocessing, the descriptive statistics of each variable are shown in

Table 2. Descriptive Statistics of Variables.

Variable Type	Variable	Observations	Mean	Std. Dev.	Min	Max
Dependent Variable	Er	3640	−0.184	0.076	−0.351	0.226
Core Explanatory Variable	Policy	3640	0.181	0.385	0	1
Control Variables	Ed	3640	10.74	0.699	9.305	12.51
	Ope	3640	0.002	0.003	0.000	0.011
	Cl	3640	0.397	0.209	0.108	0.976
	Fis	3640	0.451	0.217	0.099	0.995
	Str	3640	0.424	0.101	0.218	0.716
Mechanism Variables	Con	3640	2.054	1.236	0.389	5.843
	Tc	3624	0.148	0.210	0.001	1.782
	Pc	3640	0.020	0.025	0.001	0.119
	Di	3640	0.185	0.142	0	0.564

. Meanwhile, trend charts of economic resilience and mean differences between the treatment group and the control group are plotted (Figure 2). The results show that the economic resilience of both the treatment group and the control group shows an upward trend, but the increase in the treatment group is more significant, while the control group is relatively flat; after the implementation of the information consumption pilot policy in 2013, the gap (mean difference) in economic resilience between the two groups gradually turned from negative to positive and continued to expand, initially indicating that information consumption may have a resilience-enhancing effect on the economy.

4. Empirical Analysis

4.1. Baseline Regression

The results of the baseline regression are presented in

Table 3. Baseline Regression Results.

Variable	(1)	(2)
	Without Control Variables	With Control Variables
Policy	0.084 ***	0.084 ***
	(0.01)	(0.01)
Cl		−0.088 ***
		(0.02)
Str		−0.029
		(0.02)
Ope		−4.332 ***
		(0.89)
Ed		0.045 ***
		(0.01)
Fis		−0.024
		(0.02)
N	3640	3640
Id	Controlled	Controlled
Year	Controlled	Controlled
Controls	Controlled	Controlled
R2	0.484	0.498

Note: *** indicate significance at the 1% levels; robust standard errors are in parentheses. The same below.

. Column (1) shows the regression results without including control variables; the results indicate that the impact effect of the information consumption pilot policy on the improvement of urban economic resilience reaches 0.084, which is significant at the 1% significance level. Column (2) shows the regression results after adding control variables; it can be seen that the positive impact of the policy on urban resilience remains significant, with an impact effect value of 0.084 at the 1% significance level. Thus, Research Hypothesis H1 is initially verified.

4.2. Robustness Tests

4.2.1. Parallel Trend Test

The parallel trend assumption is a core prerequisite for the effective application of the DID model. This study uses the event study method to test the dynamic effects of policy implementation. Figure 3 presents the results of the parallel trend assumption and policy dynamic effect test; it can be seen that before the implementation of the information consumption pilot policy, there was no significant difference in the average change coefficient between the treatment group and the control group, indicating that the two groups satisfy the parallel trend assumption. Meanwhile, after the policy implementation, the estimated coefficient of the treatment group gradually becomes positive and shows a dynamic upward trend, indirectly confirming that the resilience-enhancing effect of the information consumption policy on urban economic resilience is continuously being released and gradually emerging.

4.2.2. Placebo Test

To avoid spurious regression caused by omitted unobservable variables in the baseline regression, this study further conducts a permutation test. Specifically, cities equal in number to the actual pilot cities are randomly selected from the original sample cities as the pseudo-treatment group, and the remaining cities form the pseudo-control group; the policy implementation time is randomly set to construct a purely random false policy scenario. Through 500 repeated simulations, Model (1) is used to estimate the “pseudo-policy effect” of each simulation, forming the effect distribution shown in Figure 4. The results show that the simulated pseudo-effect coefficients are concentrated around 0 and approximately normally distributed, and more than 90% of their corresponding p-values are greater than 0.1; the real policy effect (marked by the red vertical line, with a coefficient of 0.084) significantly deviates from the concentration interval of the pseudo-effect, indicating that this effect cannot be explained by random interference, which strongly verifies the causal resilience-enhancing effect of the information consumption pilot policy on urban economic resilience.

4.2.3. PSM-DID Test

The list of pilot cities for information consumption is approved in batches by the Ministry of Industry and Information Technology. The selection process is susceptible to the influence of factors such as a city’s economic development level and geographical location, leading to government self-selection bias. This makes it difficult to meet strict quasi-experimental conditions and may result in endogeneity issues such as reverse causality. To address this, this study draws on the research approach of Wang Xiuhua et al. [49] and adopts the Propensity Score Matching-Difference-in-Differences (PSM-DID) method for correction. Specifically, all control variables in the baseline regression are used as matching covariates. After calculating the urban propensity matching scores via logit regression, 1:1 nearest-neighbor matching is conducted.

As can be seen from

Table 4. PSM Validity Test.

Variable	Matched or Not	Mean Value		t-Test
		Treatment Group	Control Group
Cl	Unmatched	0.517	0.370	16.87 ***
	Matched	0.517	0.504	1.01
Str	Unmatched	0.485	0.410	18.11 ***
	Matched	0.485	0.476	1.58
Ope	Unmatched	0.003	0.002	7.87 ***
	Matched	0.003	0.003	1.01
Ed	Unmatched	11.301	10.611	24.84 ***
	Matched	11.301	11.333	−0.92
Fis	Unmatched	0.572	0.424	16.49 ***
	Matched	0.572	0.591	−1.47

Note: *** indicate significance at the 1% levels; robust standard errors are in parentheses.

(PSM Validity Test), the t-test results of the covariates after matching are all insignificant, indicating that there is no difference in covariates between the treatment group and the control group, and the PSM results are valid. Additionally, the balance test (Figure 5) shows that there was a large bias between the control group and the treatment group before matching, while after matching, the bias of each covariate is less than 10%, indicating good matching quality, which is consistent with the conclusions in Table 4.

On this basis, further Difference-in-Differences estimation is carried out. The results in

Table 5. PSM-DID Regression Results.

Variable	Nearest-Neighbor Matching (1:1)
Policy	0.041 ***
	(0.01)
N	968
Id	Controlled
Year	Controlled
Controls	Controlled
R2	0.608

Note: *** indicate significance at the 1% levels; robust standard errors are in parentheses.

show that after effectively overcoming sample self-selection bias, the policy variable (Policy) all passes the significance test, and the core empirical conclusion remains robust. That is, the implementation of the information consumption pilot policy can significantly enhance urban economic resilience, and the baseline regression results are reliable.

4.2.4. Heterogeneous Treatment Effect

The traditional Two-Way Fixed Effects (TWFE) estimation framework has a critical limitation. When there is cohort heterogeneity or time heterogeneity in the treatment effect, this method may produce biased estimation. Sun and Abraham [50] proposed an estimation method based on interactive weights to calculate the average treatment effect for different periods and groups, which is applicable when there is sufficient sample size, there are never-treated individuals, and there is no policy exit. This section adopts the S-A statistic to conduct robust estimation for heterogeneity. The specific implementation paths include two aspects: first, calculate the conditional average treatment effect under the combination of different groups and periods, and then obtain the estimated value of the Average Treatment Effect on the Treated (ATT) through weighted aggregation; second, plot a dynamic treatment effect trend chart with a 95% confidence interval based on the S-A estimation results. As shown in

Table 6. Average Treatment Effect.

Variable	ATT
Policy	0.089 ***

Note: *** indicate significance at the 1% levels; robust standard errors are in parentheses.

and Figure 6, after eliminating heterogeneity interference with the S-A statistic, the average treatment effect (ATT) of the core explanatory variable is still significantly positive, and the trend characteristics of the dynamic effect chart are basically consistent with the benchmark regression results, which verifies the robustness of the conclusion of this paper.

4.2.5. Other Robustness Tests

(1)

Replacing the Measurement Method of the Dependent Variable

To further ensure the robustness of the baseline regression results, based on the core connotation of economic resilience, this study selects 14 measurement indicators from three dimensions of urban economic resilience—namely shock resistance capacity, adaptability, and transformation and innovation capacity—to construct a comprehensive evaluation index system [51], as shown in

Table 7. Economic Resilience Index System.

Target Layer	Criterion Layer	Measurement Indicators	Attribute
Economic Resilience	Resistance Capacity	Per Capita GDP	+
		Registered Urban Unemployment Rate	-
		Upgrading of Industrial Structure	+
		Year-end Balance of Urban and Rural Residents’ Savings	+
	Adaptability Capacity	Local Fiscal Expenditure	+
		Total Retail Sales of Social Consumer Goods	+
		Year-end Deposit Balance of Financial Institutions	+
		Number of Beds in Hospitals and Health Centers	+
		Industrial Smoke and Dust Emissions	-
	Innovation Capacity	Urbanization Rate	+
		Fixed Asset Investment	+
		Number of Patent Authorizations	+
		Fiscal Expenditure on Education	+
		Fiscal Expenditure on Science and Technology	+

. The entropy weight-TOPSIS method is adopted to calculate the comprehensive index score: first, the entropy weight method is used to compute the weight of each indicator, and then the final score is obtained by calculating the distance between each indicator and the positive ideal solution. A higher score indicates stronger economic resilience.

From the scatter plot of the measurement indicators and the economic resilience from the baseline regression (Figure 7), it can be observed that there is a positive correlation between the two. The corresponding regression results are presented in Column (1) of

Table 8. Results of Other Robustness Tests.

Variable	(1)	(2)	(3)
	Replaced Dependent Variable Measurement	Excluding Municipalities	Sample Shrinkage (5% Winsorization)
Policy	0.019 ***	0.074 ***	0.070 ***
	(0.00)	(0.01)	(0.00)
N	3640	3588	3640
Id	Controlled	Controlled	Controlled
Year	Controlled	Controlled	Controlled
Controls	Controlled	Controlled	Controlled
R2	0.854	0.478	0.542

Note: *** indicate significance at the 1% levels; robust standard errors are in parentheses.

. The promotion effect of the information consumption pilot policy on urban economic resilience is significantly positive at the 1% significance level, with an effect size of 0.019. This conclusion is basically consistent with the baseline regression results, further confirming the robustness of the research conclusions.

(2)

Excluding Municipalities Directly Under the Central Government

Considering that the economic structure of municipalities directly under the central government is significantly different from that of other prefecture-level cities, which may cause estimation bias in the baseline regression results. To this end, this study excludes the four municipalities directly under the central government (Beijing, Shanghai, Tianjin, and Chongqing) from the sample and re-conducts the regression estimation; the corresponding results are reported in Column (2) of Table 8. The empirical results show that the estimated coefficient of the core explanatory variable is significantly positive at the 1% significance level, with a value of 0.074, further verifying the robustness of the research conclusions.

(3)

Sample Shrinkage (Winsorization at 5% Bilateral Level)

To mitigate estimation bias caused by sample outliers, this study conducts 5% bilateral winsorization on the core variables. The corresponding results are shown in Column (3) of Table 8; the estimated coefficient of the information consumption pilot policy variable is still significantly positive at the 1% significance level, further verifying the robustness of the research conclusions.

4.3. Mechanism Analysis

The baseline regression results have confirmed that the information consumption pilot policy has a significant resilience-enhancing effect on urban economic resilience. Based on the theoretical analysis framework above, this section further explores the internal mechanism through which the policy affects urban economic resilience.

4.3.1. Mediation Effect Test

(1)

Consumption Growth Effect: Considering that consumption behavior has inertial characteristics and there may be a bidirectional causal relationship between current consumption growth and the information consumption pilot policy, which is prone to endogeneity issues, this study incorporates the one-period lag of residents’ consumption into the regression to construct a dynamic mediation effect model. Column (1) of

Table 9. Mediation Effect Test Results.

Variable	(1)	(2)	(3)
	Consumption Growth Effect	Technological Innovation Effect	Human Capital Improvement Effect
Policy	0.386 ***	0.088 ***	0.002 ***
	(0.03)	(0.01)	(0.00)
N	3360	3348	3360
Id	Controlled	Controlled	Controlled
Year	Controlled	Controlled	Controlled
Controls	Controlled	Controlled	Controlled
R2	0.872	0.857	0.894

Note: *** indicate significance at the 1% levels; robust standard errors are in parentheses.

shows that the estimated coefficient of the information consumption pilot policy on consumption growth is 0.386, which is significantly positive at the 1% level. As the core carrier of new-type consumption, the information consumption pilot policy directly drives the expansion of consumption demand and the upgrading of consumption structure by improving digital consumption infrastructure and enriching information consumption scenarios. As the core driver of economic growth, consumption growth can further consolidate the urban economic foundation, enhance the buffering capacity of the economic system in response to external shocks, and thus become an important transmission path for improving urban economic resilience.

(2)

Technological Innovation Effect: Considering that technological innovation has an R&D cycle and a time lag effect in achievement transformation, and there may be a bidirectional causal relationship between the information consumption pilot policy and technological innovation (for example, cities with higher technological innovation levels are more likely to be included in the pilot scope), which is prone to endogeneity bias. To improve the accuracy of mediating effect identification, this study incorporates the one-period lag of technological innovation into the regression to construct a dynamic mediation effect model. Column (2) of Table 9 shows that the estimated coefficient of the information consumption pilot policy on technological innovation is 0.088, which is significantly positive at the 1% level. This result confirms the linkage mechanism between information consumption and technological innovation: the expansion of information consumption demand forces enterprises to increase investment in digital technology R&D and promote the transformation and application of technological achievements; as the core engine of economic growth, technological innovation can optimize industrial production efficiency, cultivate new growth points, enhance the adaptability and transformation capabilities of the urban economy in the face of shocks, and thus improve economic resilience through the technological innovation channel.

(3)

Human Capital Improvement Effect: Considering that human capital accumulation has progressive and persistent characteristics, labor skill improvement is not achieved instantaneously, and there may be a bidirectional causal relationship between the information consumption pilot policy and the level of human capital (for example, human capital-intensive cities are more likely to become pilots), which in turn gives rise to endogeneity issues. To more accurately identify the mediating effect, this study incorporates the one-period lag of human capital into the regression to construct a dynamic mediation effect model. Column (3) of Table 9 shows that the estimated coefficient of the information consumption pilot policy on human capital improvement is 0.002, which is significantly positive at the 1% level. With the popularization of information consumption scenarios, channels such as online education and digital skills training have significantly reduced the cost of knowledge acquisition, promoting the upgrading of the labor force’s skill structure toward digitalization and specialization. This accumulation of human capital not only improves the matching efficiency of the labor market but also enhances the adaptability of the economic system to technological changes and industrial transformation: when external shocks occur, high-skilled labor can more easily support industrial adjustment through career changes, thereby providing human support for the improvement of economic resilience.

4.3.2. Moderating Effect Test

The results of the moderating effect test are shown in

Table 10. Moderating Effect Test Results.

Variable	(1)	(2)
Policy	0.060 ***	0.054 ***
	(0.01)	(0.01)
Di	−0.035 ***	−0.001
	(0.01)	(0.00)
Policy × Di	0.244 ***	0.024 ***
	(0.04)	(0.00)
N	3640	3640
Id	Controlled	Controlled
Year	Controlled	Controlled
Controls	Controlled	Controlled
R2	0.519	0.535

Note: *** indicate significance at the 1% levels; robust standard errors are in parentheses.

: Column (1) reports the baseline test results where the level of digital infrastructure is measured by textual analysis. The coefficient of the interaction term between digital infrastructure level and the information consumption pilot policy is 0.244, which is significantly positive at the 1% significance level. This indicates that the level of digital infrastructure plays a positive moderating role in the process of the information consumption pilot policy enhancing urban economic resilience: the more improved a city’s digital infrastructure, the stronger the resilience-enhancing effect of the information consumption pilot policy. Improved digital infrastructure can provide hardware support for the implementation of information consumption scenarios and the release of consumption demand, reduce information transmission costs and transaction frictions, and thus amplify the promoting effect of the policy on economic resilience.

In addition, considering that the method of measuring digital infrastructure based on textual analysis may have subjective biases due to differences in terminology definition standards and text segmentation rules, as a core hard indicator of digital infrastructure, broadband penetration rate can not only directly and objectively reflect the stability and speed of urban information transmission as well as the basic conditions for the implementation of consumption scenarios, but also is more in line with the core demand of information consumption for high-speed network support. To further improve the reliability of the conclusions, this study uses broadband penetration rate to re-characterize the level of digital infrastructure and constructs an optimized moderating effect model. The test results are shown in Column (2) of Table 10: the coefficient of the interaction term between broadband penetration rate and the information consumption pilot policy is still significantly positive at the 1% significance level, and the sign of the coefficient is consistent with the baseline results, which further verifies the robustness of the positive moderating role of digital infrastructure.

4.4. Heterogeneity Analysis

4.4.1. Heterogeneity in Population Size

The annual end total population is used to measure the urban population size, and based on the median, the sample is divided into two groups: cities with large population size and cities with small population size for regression. Columns (1) and (2) of

Table 11. Heterogeneity Test Results.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	Large Population Size	Small Population Size	High Economic Density	Low Economic Density	Resource-Based Cities	Non-Resource-Based Cities
Policy	0.102 ***	0.029 ***	0.090 ***	0.033 ***	0.043 ***	0.096 ***
	(0.01)	(0.01)	(0.01)	(0.01)	(0.00)	(0.01)
N	1830	1808	1822	1798	1443	2197
Id	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
Year	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
Controls	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
R2	0.621	0.385	0.561	0.501	0.483	0.519
Intergroup Coefficient of Variation Test	0.073 ***		0.057 ***		−0.053 ***

Note: *** indicate significance at the 1% levels; robust standard errors are in parentheses.

show that the estimated coefficients of the information consumption pilot policy are both significantly positive at the 1% significance level, and the inter-group difference coefficient passes the test. Among them, the promoting effect on the economic resilience of cities with large population size is significantly higher than that of cities with small population size. The possible reason is that cities with large population size have a larger consumer base, which can provide a natural scale effect for information consumption scenarios. The higher density of consumption demand brought by population agglomeration makes it easier for enterprises to spread the investment cost of information consumption infrastructure through user scale, promoting service innovation and scenario implementation; at the same time, cities with large population size have more complete digital infrastructure coverage and more sound supporting services, which can more efficiently undertake the dividends of the information consumption pilot policy, thereby stimulating a stronger driving force for economic resilience improvement.

4.4.2. Heterogeneity in Economic Density

Economic density is measured by the ratio of regional GDP to the land area of the administrative region. Based on the median, the sample is divided into two groups: cities with high economic density and cities with low economic density for regression. Columns (3) and (4) of Table 11 show that the coefficients of the information consumption pilot policy are both significantly positive at the 1% significance level, the inter-group difference coefficient passes the test, and the promoting effect on cities with high economic density is more prominent. The possible reason is that cities with high economic density have a high degree of industrial agglomeration and strong factor allocation efficiency, and the transmission path of the information consumption policy is smoother: on the one hand, high economic density means that the spatial agglomeration of enterprises, population, and technology is closer, and the technology diffusion and knowledge spillover effects brought by information consumption are more significant, which can quickly penetrate into all links of the industrial chain and promote collaborative innovation; on the other hand, cities with high economic density have a more solid digital economic foundation, and the information consumption pilot policy is more likely to form a positive cycle of “consumption upgrading → industrial transformation → resilience strengthening”, resulting in higher policy marginal benefits.

4.4.3. Heterogeneity in Resource Dependence

Taking the National Plan for the Sustainable Development of Resource-Based Cities (2013–2020) issued by the State Council as a reference, the sample is divided into two groups: resource-based cities and non-resource-based cities for regression analysis. Columns (5) and (6) of Table 11 show that the estimated coefficients of the information consumption pilot policy are both significantly positive at the 1% significance level, the inter-group difference coefficient passes the test, and the resilience-enhancing effect on non-resource-based cities is significantly stronger than that on resource-based cities.

This difference stems from the long-standing constraints faced by resource-based cities, including industrial structure lock-in, factor foundation shortages, and path dependence. Specifically, their “single-industry-dominated” industrial structure has weak compatibility with information consumption; the supply of digital infrastructure and interdisciplinary human capital is insufficient; moreover, under the “resource curse”, innovation suppression and institutional rigidity have delayed policy transmission. In contrast, non-resource-based cities, relying on their diversified industrial structure, sound digital factor foundation, and flexible market environment, can more efficiently capture the dividends of the policy, forming a positive cycle of “policy incentives → market response → resilience enhancement”.

4.5. Further Extended Analysis

Existing literature has mostly focused on analyzing the local mechanism of action of the Information Consumption Pilot Policy (ICPP), while research on whether it generates cross-regional spillover effects through spatial connections remains relatively scarce. Theoretically, the network attribute of the digital economy determines that factor mobility and technology diffusion inherently possess spatial permeability. As an institutional innovation of the digital economy in the consumption field, the ICPP, during its implementation, fosters digital consumption scenarios, derives collaborative networks of the information industry, and drives the flow of skilled human capital—all of which have the potential for spatial spillover that transcends geographical boundaries. Existing studies have confirmed that the development of the digital economy exerts significant spatial spillover effects on economic resilience and tourism economic resilience [52]. Based on this, it can be inferred that as a core consumption form of the digital economy, the implementation effect of the ICPP is likely to exhibit similar cross-regional transmission characteristics.

To verify the above theoretical inference, this study first calculates the Global Moran’s I of urban economic resilience using a spatial weight matrix (as shown in

Table 12. Global Moran’s I.

Adjacency Matrix			Economic Distance Matrix		Inverse Distance Matrix
Year	Moran’s I	p-value	Moran’s I	p-value	Moran’s I	p-value
2010	0.104	0.004	0.009	0.657	0.016	0.000
2011	0.021	0.505	0.050	0.052	0.011	0.003
2012	0.034	0.328	0.062	0.019	0.014	0.000
2013	0.046	0.194	0.075	0.005	0.014	0.000
2014	0.099	0.010	0.157	0.000	0.025	0.000
2015	0.104	0.007	0.150	0.000	0.032	0.000
2016	0.127	0.001	0.155	0.000	0.033	0.000
2017	0.124	0.001	0.174	0.000	0.033	0.000
2018	0.145	0.000	0.168	0.000	0.036	0.000
2019	0.157	0.000	0.148	0.000	0.038	0.000
2020	0.168	0.000	0.147	0.000	0.041	0.000
2021	0.160	0.000	0.158	0.000	0.036	0.000
2022	0.165	0.000	0.158	0.000	0.039	0.000

) and draws Local Moran Scatter Plots corresponding to the adjacency matrix (as shown in Figure 8, Figure 9, Figure 10 and Figure 11). The results indicate that urban economic resilience exhibits a significant spatial autocorrelation feature, with most observations in the Local Moran Scatter Plots concentrated in the first and third quadrants, showing an obvious spatial agglomeration phenomenon. On this basis, this study constructs a Spatial Difference-in-Differences Durbin Model (SDID) and adopts the partial differential effect decomposition technique to avoid the endogeneity bias of point estimation in traditional spatial econometric models, thereby accurately identifying the direct effect of the Information Consumption Pilot Policy (ICPP) on local economic resilience and the indirect spillover effect on surrounding regions.

As presented in

Table 13. Estimation Results of Spatial DID Durbin Model (SDID).

Variable	(1)	(2)	(3)
	Adjacency Matrix	Inverse Distance Weight Matrix	Economic Distance Weight Matrix
Policy	0.084 ***	0.087 ***	0.065 ***
	(0.00)	(0.00)	(0.00)
W × Policy	0.027 ***	0.523 ***	0.066 ***
	(0.01)	(0.09)	(0.01)
Direct Effect	0.086 ***	0.097 ***	0.070 ***
	(0.00)	(0.01)	(0.00)
Indirect Effect	0.055 ***	2.757 ***	0.130 ***
	(0.01)	(0.92)	(0.02)
Total Effect	0.141 ***	2.854 ***	0.200 ***
	(0.01)	(0.92)	(0.02)
ρ	0.218 ***	0.776 ***	0.341 ***
	(0.02)	(0.06)	(0.03)
N	3640	3640	3640
Id	Controlled	Controlled	Controlled
Year	Controlled	Controlled	Controlled
Controls	Controlled	Controlled	Controlled
R2	0.230	0.109	0.259

Note: *** indicate significance at the 1% levels; robust standard errors are in parentheses.

, the estimation results based on three settings—spatial adjacency matrix, inverse distance weight matrix, and economic distance weight matrix—all show the following: the spatial autoregressive coefficient (ρ) is significantly positive at the statistical level, indicating that urban economic resilience presents a significant spatial agglomeration feature, which provides a realistic foundation for the spatial spillover effect of the ICPP; the coefficient of the direct policy effect is significantly positive, further verifying the promoting effect of the ICPP on local economic resilience; the coefficient of the indirect effect is also significantly positive, suggesting that the policy can not only promote the improvement of local economic resilience but also drive the enhancement of economic resilience in surrounding cities through geographical and economic connection channels, effectively expanding the spatial radiation boundary of the policy effect.

To further reveal the differentiated characteristics of the policy’s spatial spillover effects, this study incorporates the perspective of spatial heterogeneity analysis, draws on the research framework of Cao Qingfeng [53], and constructs a grouped estimation model based on spatial distance thresholds to empirically examine the heterogeneity of the spatial spillover effects of the Information Consumption Pilot Policy (ICPP). The results are shown in Figure 12: the spatial spillover effects of the ICPP exhibit significant geographical heterogeneity, with its inherent logic rooted in the dynamic reshaping mechanism of distance on the efficiency of information resource allocation, the evolution of regional competition and cooperation relations, and the effectiveness of policy transmission. Within the 0–50 km interval of resource homogeneous competition, the information resource endowments and market structures of pilot cities and surrounding regions are highly convergent, and regional competitive relations are significantly stronger than synergistic effects, leading to the failure of the policy’s positive spillover effects to be effectively unleashed. Within the 50–300 km interval, geographical distance not only effectively mitigates the pressure of short-distance homogeneous competition but also forms positive transmission through channels such as inter-regional information industry collaboration and consumer market linkage. Moreover, as the agglomeration effect and transaction cost reach an optimal equilibrium, the measured spillover intensity peaks at around 250 km. Within the 300–550 km interval, when the distance exceeds the effective radiation range of the policy, the regional competition and cooperation pattern undergoes a structural transformation, and pilot cities gradually form a “siphon effect” on cross-regional high-quality resources (such as leading digital enterprises and consumer traffic), thereby inhibiting the improvement of economic resilience in surrounding regions. When the distance exceeds 550 km, the continuous expansion of geographical distance leads to a significant decline in policy transmission efficiency (such as information diffusion rate and industrial linkage frequency), and under the influence of the law of spatial attenuation, the spillover effects gradually converge to a statistically insignificant level. Overall, the aforementioned phased evolutionary characteristic of “resource homogeneous competition → positive synergistic driving → cross-regional resource siphon → spatial attenuation of effect” is highly consistent with the theory of regional economic competition and cooperation as well as the law of spatial attenuation, and further provides empirical evidence for the regional coordinated optimization of information consumption policies.

5. Conclusions and Implications

5.1. Conclusions

Based on panel data of 280 prefecture-level cities in China from 2010 to 2022, this study treats the information consumption pilot policy as a quasi-natural experiment and employs a multi-period Difference-in-Differences (DID) method to systematically examine the impact of information consumption on urban economic resilience and its underlying mechanisms. The key findings are as follows:

First, the information consumption pilot policy significantly enhances urban economic resilience, with a policy effect coefficient of 0.084. This conclusion remains robust after a series of robustness tests, including parallel trend test, placebo test, PSM-DID correction, replacement of dependent variable measurement, exclusion of municipalities directly under the central government, and sample winsorization.

Second, mechanism analysis reveals that the policy indirectly strengthens urban economic resilience through three transmission paths: consumption growth, technological innovation, and human capital improvement. Specifically, the policy stimulates the expansion of information consumption scale and upgrading of consumption structure to consolidate the economic foundation; drives targeted R&D in digital technology fields to enhance industrial risk resistance; and promotes the accumulation of high-skilled talents to support industrial adjustment and recovery.

Third, the level of digital infrastructure plays a positive moderating role in the above process. More complete digital infrastructure breaks technical bottlenecks, amplifies network externalities, and reduces transaction costs, thereby magnifying the resilience-enhancing effect of the information consumption policy.

Fourth, the information consumption pilot policy exhibits significant heterogeneous effects. Cities with larger populations benefit more notably from the natural scale effect of consumer groups and improved digital infrastructure; those with higher economic density leverage high industrial agglomeration and efficient factor allocation to smooth policy transmission, achieving more prominent marginal benefits; non-resource-based cities, supported by diversified industrial structures and flexible market environments, capture policy dividends more efficiently than resource-based cities constrained by industrial lock-in and path dependence, resulting in a stronger resilience enhancement effect.

Fifth, extended studies have confirmed that the pilot policies for information consumption exert a spatial spillover effect on urban economic resilience. Furthermore, this spillover effect exhibits a phased evolutionary characteristic with changes in geographical distance, specifically following the sequence of “homogeneous resource competition → positive synergistic driving → cross-regional resource siphoning → spatial attenuation of effects”.

5.2. Implications

Based on the above conclusions, this study proposes the following policy implications:

First, deepen the promotion of the information consumption pilot policy to strengthen the foundational support for urban economic resilience. Expand the scope of pilot cities, focus on improving digital consumption infrastructure (e.g., 5G networks, cloud computing centers) and innovating consumption scenarios (e.g., smart healthcare, industrial Internet services), and fully release the potential of rigid information consumption demand. This will give full play to its role as an economic stabilizer and enhance cities’ buffering capacity against external shocks.

Second, consolidate digital infrastructure construction to optimize the transmission environment of policy effects. Local governments should increase investment in new digital infrastructure, improve network coverage quality and data circulation efficiency, and eliminate the digital divide. For small and medium-sized cities, targeted measures such as fiscal subsidies and policy guidance should be adopted to make up for infrastructure shortcomings, providing hardware support for the implementation of information consumption policies and fully releasing their role in promoting economic resilience.

Third, activate mechanism transmission momentum to strengthen the coordinated empowerment of “consumption-innovation-human capital”. On the consumption side, use subsidies and standardization to expand information consumption scale and drive the coordinated development of upstream and downstream industrial chains. On the innovation side, establish special funds to support the R&D and transformation of digital technologies, and promote the penetration of innovation achievements into traditional industries. On the human capital side, improve online education and vocational training systems to cultivate digital-skilled talents, and enhance the endogenous driving force of economic resilience through multi-dimensional coordination.

Fourth, implement differentiated policies to accurately match urban development characteristics. For cities with large populations, optimize the layout of consumption scenarios and infrastructure supporting to amplify scale effects. For cities with high economic density, leverage industrial agglomeration advantages to promote in-depth integration of information consumption and industrial upgrading, and cultivate resilience development benchmarks. Meanwhile, formulate personalized policy plans based on the resource endowments and development stages of different cities to improve the accuracy and effectiveness of policy implementation.

Fifth, based on the phased characteristics of the spatial spillover effect of the Information Consumption Pilot Policy (ICPP), break down administrative barriers to amplify the overall effect of enhancing regional economic resilience. When incorporating the information consumption policy into the sustainable development plans of urban agglomerations or metropolitan areas, it is necessary to clarify the key priorities of coordination in accordance with the distance gradient. Establish cross-regional information consumption cooperation platforms, which not only promote the sharing of experience in consumption scenario construction and joint R&D of digital technologies between pilot cities and surrounding areas (especially in the 50–300 km positive synergy interval), but also need to balance the allocation of high-quality resources between pilot cities and surrounding areas through resource coordination mechanisms to avoid the siphon effect in the 300–550 km interval. For the 50–300 km core synergy interval, focus on implementing the model of “technology output from pilot cities + industrial undertaking by surrounding cities”, unblock the channels for the cross-city flow of factors such as technology, talents, and data, fully release the spillover potential of the policy through geographical and economic connection channels, and ultimately achieve the overall improvement of regional economic resilience.