Can Technology Really End Energy Poverty? China's Experience Reveals a Troubling Development Contradiction

China's expanding digital economy is helping reduce energy poverty, but the same technological transformation is producing a less favorable outcome through industrial restructuring. While stronger connections between cities and rural communities improve energy consumption conditions, industrial integration can deepen energy deprivation, revealing conflicting outcomes within the country's modernization process.

The findings emerge from "Digital Transformation and Energy Poverty Alleviation: Heterogeneous Transmission of Sub Dimensions of Urban-Rural Integration," published in the journal Sustainability. Authored by Xiaolin Zhou, Junjie Wang, Xinyan Yao, Yawen Liu and Yanan Sun, the research examines how digital development influences energy poverty across 30 Chinese provincial-level regions between 2013 and 2023.

The analysis identifies geographical integration and employment outside agriculture as important channels through which digital development improves energy conditions. Industrial integration, however, produces an unfavorable effect, while the benefits associated with digitalization remain concentrated in energy consumption rather than extending consistently to affordability and clean energy transition.

China's digital economy is changing energy access, but not every dimension of deprivation

China's energy poverty problem increasingly involves differences in energy utilization, household purchasing capacity and access to cleaner energy. Although the country's energy infrastructure is comparatively developed, rural communities continue to experience disadvantages linked to income disparities, geographical conditions and uneven access to modern services.

Digital economic development offers several potential routes for addressing these inequalities. Improved information networks can facilitate resource allocation, expand economic opportunities and support coordination between energy suppliers and consumers. Whether such improvements translate into better household energy conditions depends partly on how economic activity and infrastructure are distributed between urban and rural areas.

To investigate these relationships, the researchers constructed separate indicators covering digital infrastructure, digital applications and digital industry development. Energy poverty was evaluated through energy consumption levels, energy utilization capacity and clean energy transition, allowing the analysis to distinguish between different forms of deprivation.

The dataset contains 330 observations from provincial statistical publications, energy and environmental yearbooks, and economic databases. The researchers used fixed-effects regressions, mediation analysis and dynamic estimation techniques to examine the relationship between digital expansion and energy poverty while accounting for selected economic and regional characteristics.

Their benchmark model produces a coefficient of −0.310 for digital economic development, indicating a statistically significant negative relationship with measured energy poverty. The association remains negative across several alternative specifications, including tests that exclude major municipalities and observations from pandemic-affected years.

Dynamic estimates suggest that the relationship may strengthen over time. The difference generalized method of moments model produces a coefficient of −0.302 and a calculated longer-term effect of −1.063. However, the corresponding system estimation does not produce a statistically significant digital economy coefficient, introducing uncertainty about the precise magnitude of the longer-term relationship.

The researchers also acknowledge difficulties in establishing causality. An instrumental-variable approach encounters a weak-instrument problem, preventing it from providing reliable causal identification. The results consequently offer evidence of a consistent relationship without conclusively establishing how much energy poverty reduction can be attributed to digital development alone.

Rural connectivity and non-agricultural jobs outperform broader integration measures

Rather than functioning as a single mechanism, geographical connections, industrial restructuring and the movement of production resources influence energy poverty in different ways.

Spatial integration produces the clearest favorable result. Its estimated indirect effect is −0.100, accounting for approximately 32.3% of the relationship examined in the mediation analysis. The researchers associate this pathway with better connections between urban and rural infrastructure and improved coordination of energy resources.

Population concentration and changing land-use patterns can influence the distribution and operating costs of energy networks. In the study's interpretation, stronger spatial integration helps extend modern energy services to rural areas while reducing geographical obstacles to infrastructure provision.

Employment outside agriculture represents another significant mechanism. Digital development can reduce the cost of obtaining employment information and facilitate movement into secondary and tertiary industries, potentially improving rural households' capacity to purchase modern energy services.

The estimated indirect effect of non-agricultural employment on the energy consumption dimension is −0.079. Its corresponding relationship with energy utilization capacity receives only marginal statistical support, suggesting that the clearest benefits concern consumption conditions rather than broader improvements in household energy affordability.

The distinction becomes more pronounced when the researchers examine factor allocation as a whole. Despite favorable findings for non-agricultural employment, the broader factor-allocation indicator does not demonstrate a statistically significant overall mediating effect.

Even the combined urban–rural integration indicator fails to produce a statistically significant mediating relationship. Favorable results from spatial integration coexist with unfavorable effects from industrial restructuring, while the aggregate measure conceals these differences.

This creates an important measurement problem for development policy. Improvements in broad indicators of urban–rural integration cannot establish that each underlying economic process contributes equally to reducing energy poverty.

Industrial integration reveals the energy costs concealed within economic upgrading

Industrial restructuring emerges as the most unexpected source of tension between digital development and energy welfare. Digital technologies can strengthen links between agricultural production, manufacturing and services, but the resulting economic changes are not uniformly associated with lower energy deprivation.

The researchers estimate a positive indirect effect of 0.063 through industrial integration, indicating that this pathway is associated with higher measured energy poverty. Further examination identifies an unfavorable indirect effect of 0.120 on the energy consumption dimension.

These results contrast with the negative indirect effects observed through spatial integration and non-agricultural employment. Industrial advancement therefore appears capable of offsetting part of the energy poverty reduction associated with digital economic expansion.

The study connects this outcome to the energy pressures accompanying industrial restructuring. Upgrading production systems can generate additional energy demand and transformation costs, potentially placing pressure on household consumption before efficiency improvements or cleaner technologies produce corresponding benefits.

The researchers distinguish these near-term pressures from the potential longer-term advantages of industrial modernization. More efficient production structures and wider adoption of clean technologies could eventually improve energy conditions, although the analysis does not establish when such improvements would materialize or whether they would outweigh earlier disadvantages.

A further concern emerges from the limited reach of the identified transmission mechanisms. Spatial integration and non-agricultural employment are primarily associated with improvements in energy consumption levels, rather than producing consistently significant gains in energy utilization capacity or clean energy transition.

The results expose a separation between economic modernization and different dimensions of household energy welfare. Greater consumption does not automatically establish improved affordability, just as expanding digital infrastructure does not guarantee faster adoption of cleaner energy systems.

For policymakers, the research supports examining industrial development alongside the capacity of energy infrastructure and household budgets to accommodate changing demand. Industrial upgrading without corresponding energy provision and affordability measures may weaken some of the benefits associated with digital expansion.

Unequal regional benefits expose the limits of technology-led energy policy

The relationship between digital development and energy poverty also changes according to regional economic characteristics. Income levels, energy intensity and reliance on natural resources influence the strength of the estimated associations, challenging uniform approaches to digital investment.

Resource-based provinces demonstrate a stronger negative relationship between digital development and energy poverty than non-resource-based regions. The researchers report an additional negative interaction coefficient of −0.370, suggesting that existing energy resources and industrial infrastructure may strengthen the benefits associated with digitalization.

Energy intensity produces another differentiated result. Regions with higher energy intensity exhibit stronger poverty-reducing associations, which the researchers link to opportunities for improving energy allocation and consumption efficiency through digital applications.

Income-related results are less straightforward. The interaction between digital development and per capita income is positive, indicating that the favorable association weakens as income rises. The researchers conclude that economically disadvantaged regions may experience greater poverty-reduction benefits, although their aggregate provincial analysis cannot establish how these effects are distributed among individual households.

These differences inform the study's recommendations for differentiated public investment. The researchers advocate prioritizing infrastructure improvements and digital support according to regional economic conditions rather than applying identical measures across provinces.

They identify urban–rural energy infrastructure connectivity as a central policy priority. Extending power networks and improving coordination between rural and urban energy systems could strengthen the spatial integration pathway associated with lower energy deprivation.

Employment policy receives comparable attention. Digital employment services, skills development and measures facilitating movement into non-agricultural activities could strengthen household purchasing capacity and support improvements in energy consumption.

To address the unfavorable industrial integration pathway, the researchers recommend combining industrial upgrading with cleaner energy supply and targeted consumer protection. Proposed measures include distributed photovoltaic systems, microgrids, differentiated subsidies and assistance for households facing rising energy costs.

However, the research does not directly evaluate these interventions. Its provincial-level data also cannot capture differences in household digital literacy, local energy resources or individual experiences of energy deprivation. The authors acknowledge that more detailed observations and stronger causal identification strategies are necessary.

Future investigations could examine household and county-level conditions, evaluate specific digital development initiatives and explore why improvements in consumption are not consistently accompanied by stronger affordability or clean energy outcomes. Regional spillovers and the urban–rural digital divide also remain insufficiently understood.

For developing economies confronting disparities between cities and rural communities, China's experience provides evidence of how different economic pathways can alter the relationship between digitalization and energy deprivation. Its relevance lies in the distinction between technologies that expand economic activity and mechanisms that improve household access to adequate energy.