Energy Efficiency Gains in China May Trap Low-Income Households in a Rebound Dilemma

When a household installs a more efficient heater, better insulation, or an energy-saving appliance, the expectation is straightforward: less energy consumed, lower emissions, and a smaller utility bill. Yet decades of economic research have warned that the story rarely ends there. Cheaper effective energy services invite people to use more of them, and the money saved often gets spent on other goods whose production also consumes energy. This phenomenon, known as the energy rebound effect, has now been examined in unprecedented detail for Chinese households, and the results reveal a pattern that should unsettle policymakers far beyond China’s borders.

A new study published in the Journal of Industrial Ecology by Guangqing Xu, Qing Zhao, Xinyu Li, and Wendan Zhang of Renmin University of China, with Zhao also affiliated with The Hong Kong Polytechnic University, dissects how household energy rebound effects and welfare changes differ across income groups, urban and rural residents, and provinces in China. The research, published on 2 September 2026, links energy efficiency improvements directly to household consumption behavior and the industrial processes that stand behind the goods families buy. Its central finding is stark: a substantial number of Chinese regions may be caught in what the authors describe as a “low income-high rebound-low welfare” trap, in which the poorest and most coal-dependent households experience the strongest rebound and the weakest welfare gains from efficiency improvements.

Understanding the rebound effect requires distinguishing its two principal channels. The direct rebound effect arises when an efficiency improvement lowers the effective price of an energy service, such as heating, cooling, or lighting, prompting households to consume more of that same service. A family with an efficient air conditioner, for instance, may keep rooms cooler for longer hours because comfort has become cheaper. The indirect rebound effect operates through the wider economy: the money saved on energy bills is redirected toward other consumption categories, from food and clothing to transportation and entertainment, each of which carries its own embodied energy footprint through industrial supply chains. When both channels are combined, the net energy savings from an efficiency improvement can be substantially eroded, and in extreme cases, efficiency gains can even backfire by increasing total energy use.

The study’s methodological architecture is designed to capture both channels with household-level precision. The authors employ quadratic almost ideal demand systems, an econometric framework rooted in the classic work of Deaton and Muellbauer on consumer demand, to model how households across different income strata reallocate their spending when effective energy prices change. This demand-system approach allows the estimation of compensating variation and consumer surplus, the standard welfare measures in microeconomic theory, which quantify how much better or worse off households become following efficiency-induced price changes. To trace the energy embodied in non-energy consumption, the researchers combine environmentally extended input-output analysis with multi-regional input-output frameworks, drawing on China’s carbon emission accounts and datasets to map the industrial emissions embedded in each consumption category. Household expenditure data come from the China Family Panel Studies, a large-scale longitudinal survey that provides the granularity needed to separate income groups and regions.

The empirical results reveal pronounced heterogeneity that flat national averages would conceal. Higher-income households and urban residents tend to exhibit stronger direct rebound effects, as do households in northern provinces. This pattern is economically intuitive once the structure of household budgets is considered: wealthier families have more discretionary spending capacity, so when energy services become effectively cheaper, they can more readily expand their consumption of heating, cooling, and appliance use. Northern provinces, where winter heating demand dominates household energy budgets and coal remains a prevalent fuel, show particularly strong direct rebound, suggesting that efficiency improvements in heating systems there trigger especially large behavioral responses. The indirect rebound effect, by contrast, varies more unevenly across groups, with consumption of residence-related goods and services contributing the largest strengthening effect, while food consumption exerts the largest weakening effect on the overall rebound.

The welfare dimension of the analysis is where the study makes its most consequential contribution. Efficiency improvements do not merely change energy consumption; they change household well-being, and they change it unequally. The researchers find that welfare gains from energy efficiency improvements differ substantially across income groups and regions, and that the distribution follows a troubling logic. Households with low disposable income and provinces dependent on coal tend to combine high rebound with low welfare gains, producing the trap the authors identify. In these settings, the behavioral response to cheaper energy services is large, meaning much of the potential energy saving evaporates, while the actual improvement in household welfare remains modest. The poorest households, in effect, absorb efficiency gains into expanded energy use without enjoying commensurate improvements in living standards.

Extending the analysis from five representative provinces to all thirty provinces covered in the study, the authors conclude that a substantial number of regions may experience this low income-high rebound-low welfare trap. This geographic extrapolation matters because it transforms the finding from a set of case studies into a systemic diagnosis. China’s provincial economies differ enormously in industrial structure, fuel mix, climate, and income levels, and the study demonstrates that these differences translate directly into divergent rebound dynamics and welfare outcomes. A national energy efficiency policy calibrated to average behavior would systematically misjudge both the energy savings and the distributional consequences in precisely the regions where energy poverty and coal dependence are most entrenched.

The implications for policy are twofold, and the authors are explicit that neither goal can be pursued in isolation. First, the findings underscore the necessity of interventions that prioritize a fair transition. If efficiency improvements deliver the smallest welfare benefits to the households that need them most, then complementary measures, such as targeted subsidies, income support, or differentiated pricing, may be required to ensure that decarbonization does not widen existing inequalities. Second, the study highlights the importance of decarbonizing industrial processes themselves. Because a large share of the indirect rebound operates through the embodied energy of non-energy consumption, the carbon intensity of the supply chains behind household goods determines how damaging redirected spending actually is. Cleaner industrial production would dampen the emissions consequences of rebound even where behavioral responses remain strong.

The study situates itself within a long intellectual lineage. The rebound effect was first formalized in the energy economics literature by Khazzoom in 1980, who analyzed the economic implications of mandated efficiency standards for household appliances, and later extended to the macroeconomic level by Saunders, whose Khazzoom-Brookes postulate suggested that economy-wide rebound could be substantial. Subsequent empirical work has documented direct rebound effects for residential electricity in contexts ranging from Catalonia to Beijing, and studies of UK households by Chitnis, Sorrell, and colleagues pioneered the decomposition of direct and indirect rebound for different socioeconomic groups. Recent analyses of Chinese urban households and of economy-wide rebound in Europe and the United States have reinforced the message that efficiency alone rarely delivers the energy savings that engineering calculations promise. The new study advances this tradition by integrating welfare measurement with rebound estimation at household level across an entire large economy, and by making the distributional dimension, rather than the aggregate energy balance, the analytical centerpiece.

For the global conversation about energy efficiency as a climate strategy, the Chinese evidence carries a warning that generalizes. Efficiency improvements remain one of the most cost-effective tools for reducing emissions, and international assessments continue to rank them among the cheapest abatement options available. But the rebound effect means that the realized savings depend on how households respond, and household responses depend on income, fuel dependence, climate, and the carbon intensity of the wider economy. Where these conditions combine unfavorably, efficiency policy can quietly underdeliver while appearing on paper to succeed. The study’s call for policies that pair efficiency improvement with fair transition support and industrial decarbonization is therefore not a caveat to be footnoted but a central design principle. As China pursues its dual goals of economic development and carbon neutrality, the households least equipped to benefit from efficiency gains may also be the ones whose behavior most erodes them, and recognizing that trap is the first step toward designing energy policy that is simultaneously effective and just.

One subtle strength of the welfare framework deserves emphasis. Compensating variation and consumer surplus are not interchangeable: compensating variation measures the income a household would need to reach its original utility level after a price change, while consumer surplus is derived from the area under the demand curve. By computing both within a quadratic almost ideal demand system, the authors guard against conclusions that hinge on a single welfare metric, a concern raised in earlier theoretical work on the microeconomics of rebound.

The choice of the China Family Panel Studies also matters for interpretation. Because the survey follows the same families over time, it captures how a given household adjusts its spending as prices and efficiency change, rather than comparing different households at one moment. That longitudinal structure is what makes it possible to link rebound behavior to welfare outcomes for the same people.

The contrasting roles of residence and food in the indirect rebound are likewise instructive. Spending redirected toward housing-related goods amplifies rebound because construction, renovation, and utilities carry heavy embodied energy, while additional food spending carries a comparatively lighter industrial footprint. This suggests that the composition of redirected spending, not merely its volume, shapes the emissions consequences of efficiency gains, and that supply-side decarbonization of the housing sector could meaningfully blunt indirect rebound.

Subject of Research: Household energy rebound effects and welfare changes across income groups and provinces in China

Article Title: Household energy rebound and welfare changes in china: implications for energy efficiency improvement

Article References: Xu, G., Zhao, Q., Li, X., & Zhang, W. (2026). Household energy rebound and welfare changes in china: implications for energy efficiency improvement. Journal of Industrial Ecology. https://doi.org/10.1007/s44498-026-00168-z

Image Credits: AI Generated

DOI: 10.1007/s44498-026-00168-z

Keywords: energy rebound effect, household energy consumption, energy efficiency, welfare analysis, China, direct rebound, indirect rebound, coal dependence, fair transition, industrial decarbonization, income inequality, input-output analysis