2.1 The impact of high-quality development of the logistics industry on rural industrial upgrading

In terms of the connotation of high-quality development, it is not quantitative growth, but the simultaneous improvement of economic quality and efficiency. Rooted in global sustainable and inclusive growth pursuits, this concept aligns with the UN Sustainable Development Goals (SDGs)—notably SDG 12 (Responsible Consumption and Production) and SDG 8 (Decent Work and Economic Growth)— emphasizes green development and long-term resilience, ensuring its global relevance beyond regional contexts. Thus, high-quality logistics development entails balancing efficiency and quality with sustainable principles, encompassing not only efficiency but also service reliability, green sustainability, and supply chain adaptability. It is reflected in improved efficiency, cost reduction, service optimization and comprehensive enhancements in technological innovation, green sustainability, and supply chain coordination [5]. Traditional rural industries, dominated by low value-added, short chain agriculture with weak resilience, undergo upgrading through technological innovation, restructuring, and brand development—transforming into high-value-added, high-tech, competitive sectors. This upgrading manifests in the integration of agriculture with processing, tourism, e-commerce(e.g., agricultural product deep-processing clusters, cultural tourism chains), enhanced technological content, (smart agriculture, digital rural industries), and improved market competitiveness, ultimately achieving a qualitative leap in rural economic development and optimizing the rural industrial structure.

As a crucial component of the modern service industry, the logistics industry is closely linked to rural industrial development. High-quality logistics services play a pivotal role in accelerating the circulation of agricultural products, expanding sales channels, and enhancing their market competitiveness. Specifically, high-quality logistics development can directly facilitate rural industrial upgrading in the following ways: (1) Enhancing the efficiency of agricultural product circulation: An efficient logistics system reduces the time from production to consumption, minimizes losses, and improves the freshness and quality of agricultural products. This strengthens their market competitiveness and promotes the branding and standardization of agricultural goods. (2) Promoting rural industrial upgrading: The development of the logistics industry stimulates the growth of related sectors, including warehousing, processing, and packaging, thereby extending the agricultural product value chain. It also fosters the integration of primary, secondary, and tertiary industries in rural areas, ultimately contributing a more comprehensive rural industrial system. (3) Optimizing resource allocation: High-quality logistics services help overcome geographical constraints, enabling the more effective allocation of resources at a broader scale. This not only facilitates the efficient utilization of internal rural resources but also attracts external capital, technology, and talent, providing strong momentum for rural industrial upgrading. (4) Strengthening risk resistance: A well-developed logistics network and information system enhance rural industries’ responsiveness to market fluctuations. By mitigating risks associated with information asymmetry and logistics inefficiencies, they help ensure the stable development of rural industries. Based on the above analysis, this study proposes Hypothesis 1:

H1: High-quality logistics development has a significant impact on rural industrial upgrading.

2.2 The mediating role of industrial structure

The level of regional industrial structure is a comprehensive indicator that reflects the proportional relationships, interconnections, and upgrading trends among various industries (primary, secondary, tertiary industries, and their subdivisions) within a specific region. By reviewing relevant literature, we find that the development of the logistics industry has a significant optimization effect on the regional industrial structure. On the one hand, as a bridge connecting production and consumption, the efficient operation of the logistics industry reduces transaction costs and accelerates market responsiveness, thus promoting the collaborative development of various industries within the region [6]. On the other hand, the development of the logistics industry is often accompanied by technological and business model innovations, which can permeate other industries and drive the upgrading and transformation of the industrial structure [7]. Additionally, some studies suggest that the development of the logistics industry can extend and expand related industrial chains, thereby fostering the diversification and upgrading of the regional economy [8]. As a core component of the modern service industry, high-quality logistics development can profoundly influence and optimize the regional industrial structure. Specifically, high-quality logistics development enhances the overall economic efficiency of the region, promotes the optimal allocation and efficient utilization of regional resources, and provides more market opportunities and development space for rural industries, guiding them toward higher value-added and more technology-intensive directions. Based on this, this study proposes Hypothesis 2:

H2: High-quality logistics development can promote rural industrial upgrading by improving the level of regional industrial structure.

2.3 The moderating effects of regional fiscal support for technological innovation and digital finance

2.4 The threshold effects of agricultural labor productivity and the urban-rural income ratio

2.4.2 The threshold effect of the urban-rural income gap.

The urban-rural income gap refers to the absolute difference or relative ratio between the per capita disposable income of rural and urban residents. It reflects the balance of economic development between urban and rural areas, and embodies the consumption capacity of rural residents as well as the potential driving force for factor flows between urban and rural regions. It is an important determinant of rural economic development. Against the backdrop of high-quality logistics development, changes in the urban-rural income gap may also exert a threshold effect on rural industrial upgrading. First, from a market demand perspective, high-quality logistics (e.g., cold chain and integrated supply chain services) requires sufficient scale demand to be sustainable. When the income gap exceeds a certain threshold, rural consumption is limited to basic needs, lacking demand for high-value-added agricultural products or diversified logistics services. This reduces the profitability and coverage of high-quality logistics, thereby limiting its ability to drive rural industrial upgrading. Conversely, when the gap is below the threshold, rural consumption capacity increases, generating “demand-pull” effects that promote the circulation of high-value-added products and facilitate rural industrial upgrading. Second, from a factor supply perspective, a large income gap induces outflows of talent and capital, leaving insufficient factors (e.g., e-commerce professionals, logistics infrastructure funds) to support industrial upgrading. Even with high-quality logistics facilities, their effectiveness is constrained without adequate supporting factors. A narrower income gap attracts factor inflows to rural areas, which, when combined with high-quality logistics, forms synergies that drive industrial upgrading.

Existing literature indicates that the urban-rural income gap is a key factor influencing the coordinated development of regional economies [17]. On one hand, narrowing the gap promotes rural economic development and social progress, enhances the consumption capability and investment willingness of rural residents, and thus provides a broader market space for the expansion of the logistics industry in rural areas [18]. On the other hand, high-quality logistics development can reduce logistics costs and improve operational efficiency, thereby enhancing the market circulation and value-added of agricultural products, and potentially alleviating the urban-rural income gap to some extent [19]. However, few studies have examined how the urban-rural income gap, as a threshold factor, influences the role of logistics industry development in rural industrial upgrading. Specifically, when the income gap is small, it indicates relatively favorable economic development and social welfare in rural areas, with stronger consumption capacity and investment willingness among rural residents. In this context, high-quality logistics development can more effectively penetrate rural areas, providing convenient and efficient logistics services that promote rural industrial upgrading. In contrast, when the income gap is large, rural areas may face resource scarcity, inadequate infrastructure, and brain drain, which constrain the expansion and deepening of the logistics industry. Even if the logistics industry achieves high-quality development overall, its impact on rural industrial upgrading may be significantly weakened due to the restrictive effect of the urban-rural income gap. Based on this, it proposes Hypothesis 6:

H6: There exists a threshold effect of the urban-rural income gap in the impact of high-quality logistics industry development on rural industrial upgrading.

The theoretical analysis framework of this paper is shown in Fig 1:

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Theoretical Framework of the Mechanism Linking High-quality Logistics Development on Rural Industrial Upgrading.

https://doi.org/10.1371/journal.pone.0344923.g001

3.1 Variable description and data sources

3.1.6 Data sources and descriptive statistics.

Due to data unavailability for Tibet, Hong Kong, Macau, and Taiwan, this study uses panel data from 30 Chinese provinces from 2011 to 2022. Data are primarily obtained from the China Statistical Yearbook, China Science and Technology Statistical Yearbook, and China High Technology Industry Statistical Yearbook. Missing data were supplemented using interpolation methods where necessary. The descriptive statistics of key variables are shown in Table 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Descriptive Statistics.

https://doi.org/10.1371/journal.pone.0344923.t001

3.2 Model design

3.2.1 Baseline model and mediating effect model.

To examine the direct effect of high-quality logistics development rural industrial upgrading, as well as the mediating role of industrial structure, the following recursive mediating model is constructed:

(1)(2)(3)

Where the dependent variable nis represents the level of rural industrial upgrading, the core independent variable lh represents high-quality logistics development, and the mediating variable is the regional industrial structure level. X represents control variables. The subscripts i and t indicate region and year, respectively, while and denote individual and time fixed effects, , , and represents the random disturbance term. Model (1) examines the direct effect of high-quality logistics development on rural industrial upgrading, while Models (2) and (3) test the mediating effect of regional industrial structure. To mitigate potential endogeneity concerns arising from the possible bidirectional relationship between logistics development (lh) and rural industrial upgrading (nis), this study employs a two-way fixed effects model that controls for both province-specific and time-specific unobserved heterogeneity. Although the model does not claim strict causal identification, it effectively reduces estimation bias caused by omitted time-invariant factors. The primary objective is to explore the interaction patterns and mechanisms between logistics development and rural industrial upgrading, rather than to establish causal direction. Future research may employ instrumental variable or dynamic panel approaches to further enhance causal inference.

Given that diagnostic tests strongly reject the null hypotheses of homoskedasticity (χ²(30) = 5479.10, p < 0.001) and no serial correlation (β = 0.844, t = 26.08, p < 0.001), we employ Driscoll-Kraay standard errors in all specifications to ensure robust inference.

3.2.2 Moderation effect model.

To test whether regional fiscal support for technological innovation and digital finance level moderate the effect of high-quality logistics development on rural industrial upgrading, the following models are constructed:

(4)(5)

Where, represent the moderating variables, namely regional fiscal support for technological innovation and digital finance level, respectively, while other variables remain consistent with the previous models.

3.2.3 Threshold effect model.

To further analyze the relationship between high-quality logistics development and rural industrial upgrading, it employs Hansen’s threshold regression model, using agricultural labor productivity (apl) and urban-rural income gap (uri) as threshold variables. If a single threshold exists, the following model is used:

(6)

Where, apl serves as the threshold variable, I (·) is an indicator function, and q represents the threshold value. When , the effect coefficient of high-quality logistics development on rural industrial upgrading is ; when the effect coefficient is ; If , a threshold effect exists; otherwise, it does not. If two thresholds exist, the model extends to the following double-threshold model:

(7)

where q₁ and q₂ are two threshold values, with q₁ < q₂. These two thresholds divide the sample into three regions. If three or more threshold values exist, this method can still be applied to construct multiple threshold models. Similarly, when the urban-rural income ratio (uri) serves as the threshold variable, the corresponding threshold model is:

(8)

4.1 Baseline regression

Using STATA 17.0 software, we conducted an F-test and Hausman test on the baseline model, determining that a two-way fixed effects model is suitable for investigating the mechanism through which high-quality logistics development affects rural industrial upgrading. Table 2 presents the baseline regression results. Column (1) shows that, without considering control variables, high-quality logistics development significantly enhances rural industrial upgrading at the 1% statistical level. Column (2) adds a set of control variables, including economic development level, rural residents’ per capita disposable income, and rural residents’ expenditure on education, culture, and entertainment. Even after accounting for these controls, high-quality logistics development continues to exhibit a significant positive effect on rural industrial upgrading. This finding confirms that high-quality logistics development plays a pivotal role in promoting rural industrial upgrading, thereby validating Hypothesis 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Baseline Regression Results.

https://doi.org/10.1371/journal.pone.0344923.t002

4.2 Mediation effect test

As highlighted in the theoretical analysis, regional industrial structure may act as a critical mediating variable in the relationship between high-quality logistics development and rural industrial upgrading. The results of the mediation effect test are presented in Table 3. Columns (2) and (3) of Table 3 report the regression results with industrial structure as the mediating variable. Column (3) shows that industrial structure is positively significant at the 1% statistical level, indicating that high-quality logistics development contributes to the enhancement of regional industrial structure. Furthermore, after incorporating the mediating variable, the coefficient of high-quality logistics development decreases compared to the baseline regression, while industrial structure remains positively significant. This finding supports that industrial structure plays a mediating role in the relationship between high-quality logistics development and rural industrial upgrading. In other words, high-quality logistics development can promote rural industrial upgrading by enhancing regional industrial structure. Thus, Hypothesis 2 is validated.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Mediation Effect Results.

https://doi.org/10.1371/journal.pone.0344923.t003

4.3 Moderating effect test

As discussed in the previous theoretical analysis, local fiscal support for technological innovation and regional digital finance levels may serve as two important moderating variables in the relationship between high-quality logistics development and rural industrial upgrading. To investigate this, we introduced these variables into the baseline model and analyzed the interaction terms between high-quality logistics development and local fiscal technological innovation support, as well as regional digital finance levels, to further explore the effect of logistics development on rural industrial upgrading. To eliminate the effect of multicollinearity, the interaction terms were mean-centered. The results of the moderating effect test are presented in Table 4. Column (1) shows the moderating effect of local fiscal technological innovation support. The results indicate that the interaction term between high-quality logistics development and local fiscal technological innovation support is significant at the 1% statistical level, and the signs of the core independent variable and the interaction term are the same. This suggests that local fiscal technological innovation support enhances the positive effect of high-quality logistics development on rural industrial upgrading, validating Hypothesis 3. Column (2) shows the moderating effect of digital finance level. The results indicate that the interaction term between high-quality logistics development and digital finance level is significant at the 1% statistical level, and the signs of the core independent variable and the interaction term are the same. This suggests that digital finance enhances the positive effect of high-quality logistics development on rural industrial upgrading, validating Hypothesis 4.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Moderating Effect Test Results.

https://doi.org/10.1371/journal.pone.0344923.t004

4.4 Threshold effect test

In this study, we employed the bootstrap method with 1000 resampling iterations to examine the threshold effect of agricultural labor productivity. The results are shown in Table 5. The p-values for the single and dual threshold models with agricultural productivity as the threshold variable are 0.0070 and 0.6170, respectively. This indicates that the single threshold for agricultural productivity passes the significance test at the 1% statistical level, while the dual threshold does not pass the significance test. The single threshold value is 1.1516, with a 95% confidence interval of [1.1178, 1.1564].

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Threshold Effect Test Results — Agricultural Labor Productivity.

https://doi.org/10.1371/journal.pone.0344923.t005

The estimation results for the single panel threshold are shown in Table 6. The regression results indicate that when agricultural labor productivity is low (apl ≤ 1.1516), the coefficient of the effect of high-quality logistics development on rural industrial upgrading is 0.1397. When agricultural labor productivity is higher (apl > 1.1516), the coefficient of the effect of high-quality logistics development on rural industrial upgrading is 0.0153. This disparity in coefficients reflects a diminishing returns pattern: regions with low agricultural labor productivity have greater potential for efficiency improvements in agricultural production and rural industrial operations, so high-quality logistics (e.g., optimizing agricultural product circulation, reducing factor mismatches) yields more substantial marginal gains for industrial upgrading. Conversely, regions with high agricultural labor productivity have already attained relatively efficient resource allocation, and thus the marginal contribution of logistics development to industrial upgrading is weakened. These findings suggest that high-quality logistics development exerts a relatively large effect on rural industrial upgrading when agricultural labor productivity is low, but a comparatively smaller effect when agricultural labor productivity is high. In other words, agricultural labor productivity exhibits a threshold effect in the relationship between high-quality logistics development and rural industrial structure upgrading, thereby validating Hypothesis 5.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Threshold Regression Results — Agricultural Labor Productivity.

https://doi.org/10.1371/journal.pone.0344923.t006

Similarly, the test for the case where the urban-rural income gap is used as the threshold variable is shown in Table 7. The p-values for the single and dual threshold models for the threshold variable are 0.0400 and 0.3890, respectively. This indicates that the single threshold for the urban-rural income gap passes the significance test at the 10% statistical level, while the dual threshold does not pass the significance test. The single threshold value is 363, with a 95% confidence interval of [360.0000, 371.0000].

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. Threshold Effect Test Results — Urban-Rural Income Gap.

https://doi.org/10.1371/journal.pone.0344923.t007

The estimation results for the single panel threshold are shown in Table 8. The regression results indicate that when the urban-rural income gap is small (uri ≤ 363), that is, when the income disparity between urban and rural areas is low, the coefficient of the effect of high-quality logistics development on rural industrial upgrading is 0.0154. When the urban-rural income gap is large (363 < uri), the coefficient rises to 0.1377. This pattern reflects the “greater marginal gains in less developed areas” phenomenon: regions with a large urban-rural income gap often experience pronounced market segmentation and resource mismatches, and high-quality logistics can serve as a “bridge” connecting rural supply with urban demand, overcoming circulation barriers and more effectively unlocking the potential for rural industrial upgrading. Consequently, the promoting effect of high-quality logistics development is stronger in these regions than in regions with a small urban-rural income gap, where urban-rural integration is more advanced. This suggests that when the urban-rural income gap is small, the effect of high-quality logistics development on rural industrial upgrading is relatively limited, whereas when the gap is large, the effect is relatively large. In other words, the urban-rural income gap exhibits a threshold effect in the relationship between high-quality logistics development and rural industrial upgrading, thus verifying Hypothesis 6.

Download:

• PNG
  larger image
• TIFF
  original image

Table 8. Threshold Regression Results — Urban-Rural Income Gap.

https://doi.org/10.1371/journal.pone.0344923.t008

4.5 Robustness test

Although the bidirectional fixed effects model with Driscoll-Kraay standard errors used in the baseline regression mitigates endogeneity concerns and addresses heteroskedasticity and serial correlation, potential issues arising from omitted variables or reverse causality may persist. Therefore, it conducts robustness test using the following approaches (Table 9): First, an alternative indicator, ptes (Pure Technical Efficiency Score under Variable Returns to Scale), is employed to replace the comprehensive logistics efficiency measure as the independent variable (column (1)). Specifically, ptes captures the pure technical efficiency of the logistics industry, excluding scale effects, and reflects whether the industry optimally utilizes existing technologies to convert inputs into outputs, making it suitable for evaluating “management and resource allocation efficiency” independently of scale. Second, we include the lagged independent variable to further address reverse causality concerns (column (2)). Third, we implement province-clustered standard errors as an alternative inference method to verify the robustness of our statistical conclusions (column (3)). Fourth, we re-estimate our baseline model using Driscoll-Kraay standard errors to confirm the consistency of our main findings. Finally, we conduct a 1% winsorization treatment on all variables to mitigate the potential influence of outliers. The results across all specifications remain consistent with our baseline findings, confirming the reliability of our empirical conclusions.

Download:

• PNG
  larger image
• TIFF
  original image

Table 9. Robustness Test Results.

https://doi.org/10.1371/journal.pone.0344923.t009

5.1 Discussion

This study constructs an empirical model and a dual-moderation effect framework to investigate the collaborative relationship between high-quality logistics development and rural industrial upgrading. First, high-quality logistics development not only directly facilitates the optimization and upgrading of the rural industrial structure(H1), but also indirectly promotes rural industrial upgrading by enhancing regional industrial structure(H2). This aligns with previous studies emphasizing the positive role of logistics development in rural economic transformation. Moreover, the mediating role of industrial structure is consistent with the perspective of Yan et al. (2024) [4], while this study further clarifies the specific mechanism at work in rural contexts. Second, improvements in local fiscal support for technological innovation and digital finance levels provide strong incentives for deeper integration between logistics and rural industries, thereby amplifying the positive effect of high-quality logistics development on rural industrial upgrading. Although prior literature has acknowledged the moderating role of policy and financial factors [21], this study explicitly quantifies the moderating effects of local fiscal support for technological innovation (H3) and digital finance (H4). Finally, while most existing research on rural logistics have focused on linear relationships, this study highlights nonlinear dynamics through the threshold effects of agricultural labor productivity (H5) and the urban-rural income gap (H6). As threshold variables, agricultural labor productivity and the urban-rural income gap affect the magnitude of the effect of high-quality logistics development on rural industrial upgrading across different levels, indicating that regional heterogeneity and development stage characteristics should be fully considered when promoting the integration of logistics and rural industries.

5.2 Conclusion

This study extends the theoretical framework of rural industrial upgrading by integrating multiple effect paths, including direct, mediating, moderating, and threshold effects, and clarifies how factors like agricultural labor productivity and the urban-rural income gap constrain the effect of logistics. The mediating role of industrial structure highlights the unique urban-rural integration challenges in developing economies, offering valuable insights for comparable regions (such as Southeast Asia and sub-Saharan Africa). The findings also provide practical implications for policy-making. For government, priority should be given to supporting fiscal subsidies for rural logistics research and development and enhancing digital financial infrastructure, such as subsidizing smart warehousing and promoting mobile payment systems in rural areas. The observed threshold effects underscore the need for “phased” policies: in regions with low agricultural productivity, emphasis should be placed on fundamental logistics infrastructure (such as road network construction); whereas in regions with a large urban-rural income gap, logistics development should focus on connecting urban and rural markets. For enterprises, investment strategies should align with local conditions. In low-productivity regions, bulk transportation (such as grain and raw materials) offers higher returns, while in regions with a large urban-rural income gap, e-commerce logistics (such as rural terminal distribution) can unlock latent market potential.

Based on these findings, a three-tier progressive policy system: First, at the infrastructure level, an “intelligent logistics Hub+” action plan should be implemented, constructing national-level cold chain logistics bases for agricultural products to expand digital logistics coverage in county areas. Second, at the factor allocation level, a combination tool of “technology innovation vouchers digital inclusive finance” should be developed, establishing a rural revitalization technology special fund and promote blockchain-based agricultural product supply chain finance products. Lastly, at the institutional innovation level, a “differentiated evaluation-dynamic compensation” mechanism should be established, with special transfer payments for logistics infrastructure construction in regions with low agricultural labor productivity and pilot logistics revenue reinvestment mechanisms in regions with larger urban-rural income gaps. Critically, a collaborative governance platform should be built linking agricultural and rural authorities with financial regulators, integrating digital finance infrastructure into the development and evaluation of new agricultural business entities, thereby creating a sustainable ecosystem of “technology penetration-capital circulation-industry value addition.”

Similar evidence is found in other emerging economies. For example, in India, Negi and Negi (2022) examined humanitarian logistics challenges in disaster relief and proposed a framework for effectively managing supply chain operations, demonstrating that well-organized logistics enhances coordination, efficiency, and resilience in rural and vulnerable areas [22]. These findings support our conclusion that logistics development plays a critical role in promoting rural industrial upgrading, although the dual moderating effects of technological innovation and digital finance observed in China remain largely underexplored in other contexts.

5.3 Limitations and future directions

This study has several limitations. First, data from rural areas in China may not be fully generalizable to countries with mature rural economies, such as those in the European Union or the United States. Second, the relatively coarse granularity of rural micro-data may underestimate heterogeneous effects across regions or subpopulations. Additionally, the possible omission of subtle dynamics (e.g., how small farmers specifically benefit from logistics) may affect the accuracy of policy recommendations.

Moreover, the use of provincial-level data may obscure intra-provincial heterogeneity, including variations in infrastructure density and accessibility to digital financial services. It should be explicitly acknowledged that due to the observational nature of the data and the lack of a valid instrumental variable, strict causality cannot be fully confirmed, and the results of this study should be interpreted as robust associations. Future research could address these limitations by employing more granular datasets or conducting regional case studies, thereby enabling a deeper analysis of the relationship between logistics development and rural industrial upgrading, providing more targeted guidance for global rural revitalization efforts.