In recent years, the rapid advancement of technology and supportive national policies have positioned the low altitude economy as a pivotal force in driving industrial transformation and upgrading. As an emerging economic form, the low altitude economy encompasses activities conducted using manned aircraft and unmanned aerial vehicles (UAVs) in domains such as transportation, logistics, tourism, and emergency response. This sector is increasingly recognized as a strategic emerging industry with the potential to foster new quality productive forces and stimulate economic growth. The integration of the low altitude economy into national development plans underscores its significance in shaping modern industrial systems. This paper aims to analyze how the low altitude economy influences industrial structure upgrading through mechanisms like consumption demand enhancement and technological innovation, leveraging empirical data and spatial econometric models to provide insights for policy formulation.
The concept of the low altitude economy has evolved from traditional aviation sectors, emphasizing the utilization of low-altitude airspace resources for diverse economic activities. It represents a composite economic形态 that integrates manufacturing, services, and technology, driven by innovations in UAV systems, electric vertical take-off and landing (eVTOL) aircraft, and intelligent airspace management platforms. The development of the low altitude economy is characterized by high technological intensity, efficiency, and quality, making it a catalyst for industrial upgrading. By creating new consumption scenarios and fostering technological advancements, the low altitude economy facilitates the transition from traditional industries to knowledge-intensive and service-oriented structures, thereby enhancing overall economic resilience and competitiveness.
Existing literature highlights the multifaceted nature of the low altitude economy, with studies focusing on its definition, economic implications, and role in fostering new productive forces. Scholars have explored its potential to empower green development, stimulate innovation, and serve as a competitive frontier in global industrial landscapes. The low altitude economy is often linked to the development of low-altitude intelligent networks and infrastructure, which provide the foundation for scalable applications. However, comprehensive empirical analyses on its impact on industrial structure upgrading remain limited, particularly in terms of regional heterogeneity and spatial spillover effects. This paper addresses these gaps by constructing a robust evaluation system and employing panel data models to examine the relationship between the low altitude economy and industrial structure upgrading.
The theoretical framework posits that the low altitude economy promotes industrial structure upgrading through two primary pathways: consumption demand upgrading and technological innovation. Firstly, the low altitude economy generates novel consumption scenarios, such as urban air mobility, drone-based delivery, and aerial tourism, which cater to evolving consumer preferences for convenience, efficiency, and personalized experiences. These scenarios reshape traditional consumption patterns, driving demand for high-quality services and products. This, in turn, accelerates the integration of manufacturing and services, pushing industries toward higher value-added segments. For instance, the proliferation of low-altitude logistics and tourism services stimulates related sectors like software development, platform operations, and advanced manufacturing, fostering a shift from low-end supply to high-quality supply chains.
Secondly, technological innovation serves as the core driver of the low altitude economy, enabling breakthroughs in key areas like flight control systems, navigation, composite materials, and energy applications. The low altitude economy fosters a dynamic innovation ecosystem by facilitating collaboration among enterprises, universities, and research institutions. This leads to increased technology market transactions and the commercialization of research outcomes, which enhance the technological sophistication of traditional industries. The adoption of smart flying platforms and airspace management systems, for example, improves operational efficiency and creates new business models, such as “data + intelligence + platform” integrations. These innovations not only upgrade the low altitude economy itself but also spur cross-industry convergence, such as “low altitude + logistics” and “low altitude + environmental protection,” generating new growth poles and industrial clusters.
To quantify the development level of the low altitude economy, we establish a comprehensive evaluation index system based on three dimensions: research and development (R&D) input, industrial scale, and application environment. The指标体系 includes 16 indicators, as summarized in Table 1. R&D input captures the resources allocated to innovation, including R&D personnel, funding, education expenditure, and scientific budget. Industrial scale reflects the market size and growth potential, measured by the number of UAV enterprises, patent holdings, and related output. Application environment assesses the infrastructure and technological support, such as communication networks, artificial intelligence (AI) levels, and software services.
| Target Layer | Criterion Layer | Indicator Layer | Calculation Method |
|---|---|---|---|
| Low Altitude Economy | R&D Input | R&D Personnel Input | R&D Personnel / Employed Population (X1) |
| R&D Funding Input | R&D Expenditure / GDP (X2) | ||
| Education Funding Input | Education Fiscal Expenditure / Total Fiscal Expenditure (X3) | ||
| Science and Technology Budget Input | Science and Technology Public Budget / Total Fiscal Expenditure (X4) | ||
| Scientific and Technical Personnel | Scientific and Technical Services Employees / Employed Population (X5) | ||
| Information Service Industry Personnel | Information Service Employees / Employed Population (X6) | ||
| Industrial Scale | Number of UAV Enterprises | Number of UAV and Related Industrial Chain Enterprises (X7) | |
| Number of Enterprises with UAV Patents | Number of Enterprises Holding UAV-Related Patents (X8) | ||
| Related Output Scale | Low Altitude Economy-Related Output / GDP (X9) | ||
| Proportion of Air Transport Employees | Air Transport Employees / Employed Population (X10) | ||
| Application Environment | Infrastructure Support Capacity | Number of General Aviation Airports (X11) | |
| Artificial Intelligence Level | Industrial Robot Installation Density (X12) | ||
| Communication Infrastructure Level | Number of Internet Broadband Access Ports (X13) | ||
| Internet-Related Output | Telecom Business Volume / GDP (X14) | ||
| Software Services | Software Business Revenue / GDP (X15) | ||
| Agricultural Infrastructure Level | Total Agricultural Machinery Power (X16) |
We employ the entropy method to calculate a composite index for the low altitude economy development level, ensuring objectivity in weight assignment. The entropy method determines weights based on the information content of each indicator, avoiding subjective biases. The calculation steps are as follows:
First, standardize the positive indicators using the formula:
$$ x_{ij}^* = \frac{x_{ij} – \min(x_j)}{\max(x_j) – \min(x_j)} $$
where \( i \) denotes the region, \( j \) denotes the indicator, \( \max(x_j) \) and \( \min(x_j) \) are the maximum and minimum values of indicator \( j \) across all regions and years, and \( x_{ij}^* \) is the standardized value.
Next, compute the proportion of each indicator:
$$ P_{ij} = \frac{x_{ij}^*}{\sum_{i=1}^{n} x_{ij}^*} $$
where \( n \) is the number of regions.
Then, calculate the entropy value for each indicator:
$$ e_j = -\frac{1}{\ln n} \sum_{i=1}^{n} P_{ij} \ln P_{ij}, \quad e_j \geq 0 $$
Determine the weight coefficient:
$$ w_j = \frac{1 – e_j}{\sum_{j=1}^{m} (1 – e_j)} $$
where \( m \) is the number of indicators.
Finally, compute the composite score for the low altitude economy development level:
$$ S_i = \sum_{j=1}^{m} w_j \times x_{ij}^* $$
The score \( S_i \) ranges from 0 to 1, with higher values indicating better development of the low altitude economy.
Using panel data from 30 Chinese provinces (excluding Tibet, Hong Kong, Macao, and Taiwan) from 2012 to 2023, we calculate the composite index. Data sources include statistical yearbooks, enterprise databases, and international organizations. Missing values are addressed using linear interpolation. The weights derived from the entropy method are presented in Table 2, showing that indicators like the number of UAV enterprises and related output scale carry significant weight, underscoring their importance in the low altitude economy.
| Criterion Layer | Indicator Proxy Variable | Indicator Weight (%) |
|---|---|---|
| R&D Input | X1 | 4.707 |
| X2 | 2.831 | |
| X3 | 1.447 | |
| X4 | 5.009 | |
| X5 | 3.880 | |
| X6 | 6.879 | |
| Industrial Scale | X7 | 13.366 |
| X8 | 8.081 | |
| X9 | 10.765 | |
| X10 | 6.477 | |
| Application Environment | X11 | 3.557 |
| X12 | 1.208 | |
| X13 | 4.661 | |
| X14 | 8.546 | |
| X15 | 9.902 | |
| X16 | 8.686 |
Descriptive statistics of the composite index (Table 3) reveal a steady increase in the mean value from 2012 to 2023, indicating progressive development of the low altitude economy. The standard deviation also rises, suggesting growing regional disparities.
| Year | Number of Observations | Mean | Standard Deviation | Maximum | Minimum |
|---|---|---|---|---|---|
| 2012 | 30 | 0.106 | 0.051 | 0.286 | 0.042 |
| 2013 | 30 | 0.114 | 0.053 | 0.300 | 0.046 |
| 2014 | 30 | 0.123 | 0.055 | 0.313 | 0.053 |
| 2015 | 30 | 0.136 | 0.060 | 0.318 | 0.063 |
| 2016 | 30 | 0.145 | 0.068 | 0.341 | 0.063 |
| 2017 | 30 | 0.159 | 0.072 | 0.384 | 0.079 |
| 2018 | 30 | 0.176 | 0.073 | 0.420 | 0.105 |
| 2019 | 30 | 0.196 | 0.074 | 0.431 | 0.111 |
| 2020 | 30 | 0.209 | 0.069 | 0.407 | 0.129 |
| 2021 | 30 | 0.165 | 0.075 | 0.393 | 0.052 |
| 2022 | 30 | 0.167 | 0.077 | 0.416 | 0.050 |
| 2023 | 30 | 0.175 | 0.081 | 0.429 | 0.049 |
For empirical analysis, we specify a fixed-effects panel data model to examine the impact of the low altitude economy on industrial structure upgrading:
$$ \ln ind_{it} = \alpha_0 + \alpha_1 \ln lae_{it} + X\beta + \varphi_i + \omega_t + \varepsilon_{it} $$
where \( ind_{it} \) represents the industrial structure upgrading level of region \( i \) in year \( t \), measured by the industrial structure advancedization index:
$$ ind_{it} = \sum_{n=1}^{3} y_{int} \times n $$
Here, \( y_{int} \) denotes the proportion of the \( n \)-th industry in GDP. Alternatively, we use the ratio of tertiary to secondary industry output (\( AIS_{it} = Y_{i3t} / Y_{i2t} \)) for robustness checks. \( lae_{it} \) is the low altitude economy development index, and \( X \) includes control variables: economic development level (per capita GDP, \( z1 \)), urbanization level (urban population share, \( z2 \)), policy intervention intensity (fiscal expenditure to GDP ratio, \( z3 \)), foreign investment (actual utilized foreign capital to GDP ratio, \( z4 \)), openness (total import-export to GDP ratio, \( z5 \)), and human resources (college student share, \( z6 \)). \( \varphi_i \) and \( \omega_t \) are individual and time fixed effects, respectively, and \( \varepsilon_{it} \) is the error term.
Mechanism variables include technological innovation capability (\( creat \)), measured by technology market transaction value to GDP ratio, and resident consumption level (\( purch \)), represented by the natural logarithm of per capita consumption expenditure. Table 4 provides descriptive statistics for all variables.
| Variable | Number of Observations | Mean | Standard Deviation | Minimum | Maximum |
|---|---|---|---|---|---|
| ind | 360 | 2.41 | 0.121 | 2.132 | 2.846 |
| lae | 360 | 0.156 | 0.074 | 0.042 | 0.431 |
| z1 | 360 | 6.291 | 3.237 | 1.88 | 20.019 |
| z2 | 360 | 0.623 | 0.116 | 0.363 | 0.896 |
| z3 | 360 | 0.258 | 0.110 | 0.105 | 0.758 |
| z4 | 360 | 0.019 | 0.020 | 0 | 0.121 |
| z5 | 360 | 0.130 | 0.171 | 0.004 | 1.156 |
| z6 | 360 | 0.022 | 0.006 | 0.009 | 0.044 |
| purch | 360 | 0.021 | 0.032 | 0 | 0.195 |
| creat | 360 | 9.793 | 0.379 | 8.888 | 10.869 |
Baseline regression results (Table 5) demonstrate that the low altitude economy significantly promotes industrial structure upgrading. In all model specifications, the coefficient for \( \ln lae \) is positive and statistically significant at the 1% level. Specifically, a 1% increase in the low altitude economy development level leads to a 0.026% rise in the industrial structure advancedization index. This confirms the positive role of the low altitude economy in driving industrial transformation.
| Variable | (1) Pooled Regression | (2) Pooled Regression | (3) Fixed Effects | (4) Fixed Effects |
|---|---|---|---|---|
| Low Altitude Economy | 0.066*** (14.7057) | 0.016*** (4.0156) | 0.025*** (5.3443) | 0.026*** (5.6041) |
| Constant | 1.008*** (111.6768) | 0.805*** (56.0116) | 1.010*** (104.8105) | 1.082*** (19.4526) |
| Control Variables | No | Yes | No | Yes |
| Time Fixed Effects | No | No | Yes | Yes |
| Individual Fixed Effects | No | No | Yes | Yes |
| Number of Observations | 360 | 360 | 360 | 360 |
| R² | 0.375 | 0.781 | 0.962 | 0.973 |
Robustness checks are conducted by replacing the dependent variable with the alternative industrial structure advancedization index (\( AIS_{it} \)) and excluding COVID-19 affected years (2020-2022). As shown in Tables 6 and 7, the results remain consistent, affirming the robustness of our findings.
| Variable | (1) | (2) |
|---|---|---|
| Low Altitude Economy | 0.172*** (3.0846) | 0.199*** (4.1614) |
| Constant | 0.917*** (7.6671) | 2.303*** (4.2092) |
| Control Variables | No | Yes |
| Time Fixed Effects | Yes | Yes |
| Individual Fixed Effects | Yes | Yes |
| Number of Observations | 360 | 360 |
| R² | 0.937 | 0.959 |
| Variable | (1) | (2) |
|---|---|---|
| Low Altitude Economy | 0.020*** (2.8127) | 0.024*** (3.9059) |
| Constant | 0.996*** (73.9409) | 1.089*** (14.2811) |
| Control Variables | No | Yes |
| Time Fixed Effects | Yes | Yes |
| Individual Fixed Effects | Yes | Yes |
| Number of Observations | 270 | 270 |
| R² | 0.965 | 0.975 |
Mechanism tests validate the pathways of consumption and technological innovation. Table 8 shows that the low altitude economy positively influences resident consumption, which in turn promotes industrial structure upgrading. Similarly, Table 9 confirms that technological innovation acts as a mediator, with the low altitude economy boosting innovation activities that drive structural changes.
| Variable | (1) Consumption | (2) Consumption | (3) ln ind | (4) ln ind |
|---|---|---|---|---|
| Low Altitude Economy | 0.178*** (9.4176) | 0.103*** (4.7913) | ||
| Consumption | 0.069*** (4.7901) | 0.080*** (5.8975) | ||
| Constant | 10.500*** (233.4249) | 8.826*** (34.0871) | 0.263* (1.8023) | 0.243** (2.0561) |
| Control Variables | No | Yes | No | Yes |
| Time Fixed Effects | Yes | Yes | Yes | Yes |
| Individual Fixed Effects | Yes | Yes | Yes | Yes |
| Number of Observations | 360 | 360 | 360 | 360 |
| R² | 0.987 | 0.991 | 0.962 | 0.973 |
| Variable | (1) Technological Innovation | (2) Technological Innovation | (3) ln ind | (4) ln ind |
|---|---|---|---|---|
| Low Altitude Economy | 0.028*** (5.3262) | 0.040*** (7.1057) | ||
| Technological Innovation | 0.249*** (4.1262) | 0.241*** (4.0512) | ||
| Constant | 0.086*** (7.1675) | 0.320*** (4.9615) | 0.952*** (299.0039) | 0.880*** (19.4526) |
| Control Variables | No | Yes | No | Yes |
| Time Fixed Effects | Yes | Yes | Yes | Yes |
| Individual Fixed Effects | Yes | Yes | Yes | Yes |
| Number of Observations | 360 | 360 | 360 | 360 |
| R² | 0.902 | 0.926 | 0.960 | 0.971 |
Heterogeneity analysis examines regional variations, dividing samples into northern and southern regions based on the Qinling-Huaihe line, and into high- and low-urbanization areas. Tables 10 and 11 indicate that the low altitude economy has a stronger promoting effect in northern regions and areas with lower urbanization levels, likely due to greater potential for structural transformation and resource availability.
| Variable | (1) Northern Regions | (2) Southern Regions |
|---|---|---|
| Low Altitude Economy | 0.025*** (3.7227) | 0.012** (2.3329) |
| Constant | 1.103*** (11.8448) | 0.991*** (14.2547) |
| Control Variables | Yes | Yes |
| Time Fixed Effects | Yes | Yes |
| Individual Fixed Effects | Yes | Yes |
| Number of Observations | 192 | 168 |
| R² | 0.975 | 0.976 |
| Variable | (1) High Urbanization | (2) Low Urbanization |
|---|---|---|
| Low Altitude Economy | 0.016* (1.8231) | 0.026*** (3.9024) |
| Constant | 1.083*** (16.1359) | 0.882*** (12.8872) |
| Control Variables | Yes | Yes |
| Time Fixed Effects | Yes | Yes |
| Individual Fixed Effects | Yes | Yes |
| Number of Observations | 132 | 228 |
| R² | 0.990 | 0.864 |
Threshold effect analysis explores nonlinear relationships using panel threshold models. Marketization degree (ae), measured by the number of enterprises per capita, and government support intensity (gov), represented by the fiscal expenditure to GDP ratio, serve as threshold variables. Bootstrap tests (Table 12) confirm a single threshold for marketization and a double threshold for government support. The results (Table 13) reveal that the impact of the low altitude economy intensifies with higher marketization but diminishes with excessive government support, indicating conditional effects.
| Threshold Variable | Model Mechanism | F-value | P-value | Threshold Value |
|---|---|---|---|---|
| ae | Single Threshold | 33.64 | 0.0067 | 1.0926 |
| Double Threshold | 12.90 | 0.3433 | / | |
| gov | Single Threshold | 36.61 | 0.0300 | 0.2866 |
| Double Threshold | 44.35 | 0.0000 | 0.3811 | |
| Triple Threshold | 12.14 | 0.7333 | / |
| Threshold Variable | (1) | (2) |
|---|---|---|
| lae (ae ≤ 1.0926) | 0.0146*** (4.81) | |
| lae (ae > 1.0926) | 0.0249*** (7.25) | |
| lae (gov ≤ 0.2866) | 0.0313*** (9.32) | |
| lae (0.2866 < gov ≤ 0.3811) | 0.0234*** (6.89) | |
| lae (gov > 0.3811) | 0.0155*** (4.58) | |
| Constant | 0.7137*** (31.70) | 0.7628*** (35.29) |
| Control Variables | Yes | Yes |
| Number of Observations | 360 | 360 |
| R² | 0.937 | 0.940 |
Spatial econometric analysis incorporates a spatial Durbin model (SDM) to account for spillover effects. We use a geographical distance weight matrix \( w_{ij} \), where \( w_{ij} = 1 / d_{ij} \) for \( i \neq j \), and \( d_{ij} \) is the straight-line distance between provinces. The SDM specification is:
$$ \ln ind_{it} = \varphi_0 + \rho \sum_{m=1}^{n} w_{ij} \ln ind_{it} + \beta_1 \ln lae_{it} + \theta_1 \sum_{j=1}^{n} w_{ij} \ln lae_{jt} + \beta_2 X_{it} + \theta_2 \sum_{j=1}^{n} w_{ij} X_{it} + \mu_i + \delta_t + \varepsilon_{it} $$
where \( \rho \) is the spatial autoregressive coefficient, and \( \theta_1 \) and \( \theta_2 \) are coefficients for spatially lagged variables.
Global Moran’s I tests (Table 14) confirm spatial autocorrelation in both the low altitude economy and industrial structure upgrading. Model selection tests (Table 15) validate the SDM with fixed effects. The results (Table 16) show positive direct and indirect effects, indicating that the low altitude economy not only boosts local industrial upgrading but also benefits neighboring regions through spillovers.
| Year | ln lae Moran’s I | Z-statistic | P-value | ln ind Moran’s I | Z-statistic | P-value |
|---|---|---|---|---|---|---|
| 2012 | 0.080 | 3.196 | 0.001 | 0.043 | 2.325 | 0.020 |
| 2013 | 0.084 | 3.294 | 0.001 | 0.037 | 2.160 | 0.031 |
| 2014 | 0.078 | 3.126 | 0.002 | 0.042 | 2.316 | 0.021 |
| 2015 | 0.061 | 2.638 | 0.008 | 0.053 | 2.647 | 0.008 |
| 2016 | 0.060 | 2.604 | 0.009 | 0.058 | 2.769 | 0.006 |
| 2017 | 0.040 | 2.081 | 0.037 | 0.058 | 2.777 | 0.005 |
| 2018 | 0.035 | 1.962 | 0.050 | 0.068 | 2.932 | 0.003 |
| 2019 | 0.026 | 1.961 | 0.091 | 0.075 | 3.314 | 0.001 |
| 2020 | 0.030 | 1.821 | 0.069 | 0.081 | 3.482 | 0.000 |
| 2021 | 0.104 | 3.861 | 0.000 | 0.075 | 3.250 | 0.001 |
| 2022 | 0.106 | 3.912 | 0.000 | 0.080 | 3.399 | 0.001 |
| 2023 | 0.114 | 4.150 | 0.000 | 0.084 | 3.511 | 0.000 |
| Test Type | Statistic | P-value |
|---|---|---|
| Moran’s I-error | 5.680 | 0.000 |
| LM-error | 25.409 | 0.000 |
| Robust LM-error | 7.766 | 0.000 |
| LM-lag | 21.782 | 0.000 |
| Robust LM-lag | 4.138 | 0.042 |
| LR-SDM-SAR | 42.94 | 0.000 |
| LR-SDM-SEM | 46.90 | 0.000 |
| Hausman Test | 56.49 | 0.000 |
| Variable | SDM ln ind | Direct Effect | Indirect Effect | Total Effect |
|---|---|---|---|---|
| ln lae | 0.0148*** (3.87) | 0.0155*** (3.89) | 0.0631* (1.96) | 0.0786** (2.39) |
| W × ln lae | 0.048** (2.25) | |||
| ρ | 0.1689* (1.86) | |||
| Control Variables | Yes | Yes | Yes | Yes |
| Time Fixed Effects | Yes | Yes | Yes | Yes |
| Individual Fixed Effects | Yes | Yes | Yes | Yes |
| Number of Observations | 360 | 360 | 360 | 360 |
| R² | 0.362 | 0.362 | 0.362 | 0.362 |
In conclusion, the low altitude economy significantly promotes industrial structure upgrading through consumption and innovation pathways, with variations across regions and conditions. Policy recommendations include accelerating low altitude economy development by reforming airspace management, fostering regional synergies, enhancing consumption and innovation, optimizing government support, improving marketization, and strengthening spatial coordination to maximize spillover effects. These measures can harness the full potential of the low altitude economy as a driver of industrial transformation and sustainable growth.