The nexus of energy consumption and economic performance in China

ArticleDetailDownload PDF
Many studies have explored the nexus between energy consumption and economic growth in countries across the world, although individual countries have produced different results in terms of causality and the direction of any causality. In this study, Dr Yiming He from South China Agriculture University introduces industrialisation and urbanisation as additional factors that may also evidence a nexus in the exploration of energy consumption and economic performance.

The relationship, or nexus, between energy consumption and economic performance has been the subject of many studies covering most regions of the world and, interestingly, the findings sometimes differ when countries are looked at individually. However, not many of the studies explore the nexus between the energy consumption, urbanisation, industrialisation and economic development of China.

This is where the research carried out by Dr Yiming He and his postgraduate student Meng Chen of South China Agriculture University comes in. Their aim is to illustrate the nexus between energy consumption (identifying and separating various different sources of energy), urbanisation (the almost inevitable clustering of population around the areas of growing industrialisation), the growth in industrialisation itself (creating the demand for labour that encourages urbanisation) and economic growth. The research team draw from available data from China for the period 1983 to 2014.

A review of the literature
From the existing literature, Johnson and Meuller (1973) examined how metropolitan growth affected the size and structure of (overall) consumption in Sweden, identifying that the metropolitan setting, and changes in it, affect the size and structure of consumption, and concluding that there are great and strategic regional variations in consumption patterns.

Focusing on China, Tang and Croix (1993) used province-level cross-section data to explore the relationship between energy consumption and economic activity. Their key findings include the conclusions that, when a province exports energy or has significant amounts of heavy industry, its energy consumption is higher. However, energy consumption is lower in coastal provinces than inland provinces, but the income elasticity is higher in the rapidly developing coastal provinces.

Beijing financial district at night.

From a literature review of the nexus between energy consumption and economic performance alone (that is excluding the factors of urbanisation and industrialisation), four possible distinct hypotheses were identified by the research team: Conservation Hypothesis
This suggests that there is a unidirectional causality running from economic growth and gas consumption. Quite simply, when the economy is doing well, more gas will be consumed.

Their aim is to illustrate the nexus between energy consumption, urbanisation, the growth in industrialisation itself, and economic growth.

Growth Hypothesis
This hypothesis postulates that there is unidirectional Granger causality running from gas consumption to economic growth; in other words, gas consumption is a factor in economic growth.

Feedback Hypothesis
In this case, the hypothesis suggests that there is a bidirectional causality and, in this case, economic growth and gas (power) consumption are mutually influenced.

Technicians examine pipes at a natural gas purification plant in Dazhou city, China.

Neutrality Hypothesis
Finally, this fourth hypothesis proposes that there are no direct causal links, either unidirectional or bidirectional between the two factors being tested.

However, while some of these hypotheses have been developed in relation to empirical studies that explore the nexus between energy consumption and economic growth, and some explore the nexus between economic growth and urbanisation, none have explored the causality among GDP, energy consumption, industrialisation and urbanisation in the same model. There has, therefore, been no investigation of the interactive relationship between these factors.

Dr He looked at the different types of energy consumption: electricity, gas, natural gas and liquified petroleum gas.

Drawing from the three literature reviews, Dr He and Chen provide insight into possible relationships among these four variables.

He, Fullerton, and Walke (2017) analyse the relationship between electricity consumption per capita and metropolitan economic growth for Guangzhou, China using 64 years (1949 to 2014) of annual-frequency data (although it should be noted that from 1949 to 1977, China was following a policy of collectivism and a centrally controlled economy, and from 1978 to 2014, the country had returned to a free market commercial economy).

The study applies a statistical concept known as Granger causality [Granger, 1969], that looks at two sets of time series data to see how they are related for future prediction (in this case energy consumption and economic growth). He and Gao (2017) estimate the relationships between urbanisation, industrialisation, gross electricity consumption and metropolitan economic growth with annual data from 1950 to 2013 (see the note above for the change of political policy during this period) for Guangzhou, China’s third largest urban economy after Beijing and Shanghai.

The researchers examine the relationship between increased urbanisation and industrialisation and energy consumption in China.

Dr He demonstrates that there is Granger causality from gross electricity consumption to GDP for the long run and GDP Granger causality for gross electricity consumption in the short run, while there is bidirectional Granger causality between urbanisation to gross electricity consumption and unidirectional Granger causality from industrialisation to gross electricity consumption in the short run.

The study applies a statistical concept known as Granger causality, that looks at two sets of time series data to see how they are related for future prediction.

In a further study, He and Gao (2017b) build a theoretic model to estimate the relationship between gas consumption and metropolitan economic performance with annual data from 1978 to 2013 for Guangzhou in China. Based on this shorter time period (and outside of the political policy changes) empirical results show that there is Granger causality from GDP to gas consumption for the long run in Guangzhou.

In the current paper, Dr He takes the exploration further by looking at subsets of the energy consumption data, to explore the relationships as they apply to electricity consumption, gas consumption, natural gas consumption and, finally, liquified petroleum gas (LPG) consumption. Additionally, energy consumption has been stated and used per capita in the analysis.

Conclusion and implications
The initial conclusion is to describe the nexus between energy consumption and economic performance in China, for which the Granger Causality Test has been used. The empirical results show that among four different types of energy consumption per capita (electricity consumption per capita, gas consumption per capita, natural gas consumption per capita, liquefied petroleum consumption per capita) there is some variation. For electricity consumption and gas consumption, the results are consistent with the neutrality hypothesis; in other words, there is no causality, either unidirectional or bidirectional, between those factors or economic growth. For natural gas, there is unidirectional causality from economic performance to consumption in the short run and long run, i.e. consistent with the conservation hypothesis. Finally, for LPG, there is causality between LPG consumption to economic performance in the short run, consistent with the growth hypothesis.

Clearly, the nexus between energy consumption, from all sources, and economic performance is a complex issue with further research required to create generalisable and reliable models that can help in future planning.

The period covered by the data represents two very different economic environments for the factors and variables examined; do you think the results would have been similar if you had studied each of those two policy periods separately?

The relationships between energy consumption and economic performance prior to and after 1978 reform are still uncertain.


  • Zhipeng Du and Yiming He. 2017, “Nexus between Energy Consumption and Economic Performance in China”, Journal of Business and Economics, 8(9): 1292-1305.
  • Granger, J. (1969), Investigating Causal Relations by Econometric Models and Cross-spectral Methods. Econometrica 37 (3),424-438.
  • He, Yiming, Thomas M. Fullerton, and Adam G. Walke. 2017. “Electricity Consumption and Metropolitan Economic Performance in Guangzhou: 1950–2013.” Energy Economics 63: 154–60.
  • He, Yiming, and Shaohui Gao. 2017a. “Economic Growth, Urbanization, Industrialization, and Metropolitan Electricity Consumption: Evidence from Guangzhou in China Division of Resource Management.” The Empirical Economics Letters 16 (March): 195–208.
Research Objectives
Dr Yiming He examines the relationship between energy consumption and economic performance in China.

This work was supported by National Ten Thousand Outstanding Young Scholar Program (Grant Number: W02070352) as well as Key Project of National Natural Science Foundation in China (Grant Number: 71742003)


  • Yiming He and Meng Chen

Dr He is a South China Agriculture University PhD, Professor and PhD Advisor, who won the Ten Thousand Program of the national youth talent support project, national natural science twice, national social science three times, and China Scholarship Council visiting scholar program. He has published more than one hundred papers on Man and the Economy and Energy Economics. Dr Yiming He is also the Ronald Coase Institute Young Fellow, Hong Kong Baptism University Adjunct Researcher and The University of Texas Visiting Professor.

Prof Yiming He
2760 Anise Dr,
El Paso,
TX 79936,

T: +13802415862

Related posts.


Leave a Comment

Your email address will not be published. Required fields are marked *

Share this article.

Share on facebook
Share on twitter
Share on whatsapp
Share on linkedin
Share on reddit
Share on email
Scroll to Top