Regional Electricity Demand and Economic Transition in China: A provincial panel data analysis of economic development and electricity consumption
Jiang Lin, China Energy Group, Lawrence Berkeley National Laboratory,
Xu Liu, China Energy Group, Lawrence Berkeley National Laboratory,
Gang He, Deparatment of Technology and Society, Stony Brook University,
Jin Guo, Macro Economic Research Institute of National Development and Reform Commission,
Overview
China’s economy is going through a major transition, characterized by a slower growth rate, a structural shift to the tertiary sector, and industrial deleveraging—a process to address overcapacity that has built up in key industrial sectors over the past decades. China’s Gross Domestic Product (GDP) growth rate was 7.3% in 2014, 6.9% in 2015 and 6.7% in 2016 . Contribution of the tertiary industry to total GDP has exceeded 50% since 2015 and continued to grow to 51.6% in 2016.As a consequence, demand growth for electricity is expected to slow down substantially. Electricity growth slowed to 3% in 2014 and to 0.96% in 2015, however, bounced back to 5% in 2016. The newly released National 13th Five-Year Plan For Electricity Development projected future electricity consumption over the next five years at a relatively fast pace, with a projected annual growth rate of 3.6% to 4.8% before 2020. These are still significantly higher than the average over the last three years.
Methods
We consider the following econometric model to examine the effects of current economic transition for electricity consumption in China using provincial panel data:
yit =Zitβ +ηi+εit
where yit is total electricity consumption of province i in year t;Zit is a vector of exogenous variables, including total GDP, industry composition, heavy industry capacity, and population; β is a vector of parameters; ηirepresents the individual effect, capturing the idiosyncratic characters of each province; andεitis the error term.
We first regressed electricity consumption on total GDP, tertiary share, crude steel output, and population in a linear form using ordinary least squares(OLS) in Model 1. In Model 2, we used least squares with dummy variable (LSDV) to control for the unobserved heterogeneity for each province by introducing a province dummy.Time trend (year) was also included to control for both heterogeneous variations across provinces and exogenous shocks common to all provinces in technology progress.We usedprovincial fixed effects estimators in Model3 to regress electricity consumption on total GDP, tertiary share, crude steel output, and population.
We also estimated the elasticity of economic growth on electricity consumption:
Lnyit = LnZitβ +ηi+εit
We regressed electricity consumption on GDP, crude steel production, population, and tertiary GDP share using the log-log function form for each variable in Model 4. Model 5 used LSDV to capture differences in provinces and Model 6 was estimated by the provincial fixed effects.
Results
The results show that GDP andcrude steel output have significant positive effects on total electricity consumption; however, the results for impacts of tertiary share vary. Model 2, Model 4 and Model 5 show a significant negative effect of tertiary share on electricity consumption at 95% significant level, 99% significant level, and 95% significant level , respectively. However, results of Model 1, Model 3 and Model 6 show that there is no remarkable evidence supporting the contribution to electricity savings from a transition to tertiary sector. Results from all models, except for Model 6, show a significant positive effectof population on total electricity consumption.
We also used Model 1 and Model 4 to forecast electricity consumption in 2020. Model 1 predicted electricity consumption in 2020 to be 6,712 TWh, with an annual growth rate of 3.3percent. Byassumingthat the crude steel output would stay at the same level as it was in 2015, Model 1 predictedelectricity consumption in 2020 to be 6,750 TWh, with an annual growth rate of 3.5percent. Model 4 predicted electricity consumption in 2020 to have an annual growth rate of 2.7percent, assuming a 10percent reduction in crude steel output by 2020.
Conclusions
Given the uncertainties facing China’s current economic transition, it is important to examine whether future electricity consumption will remain at a high growth or shift to a lower level in the medium term.This analysis attempts to examine the relationship between electricity consumption, and GDP, economic structure, and overcapacity in heavy industries in China, using provincial level data from 1995-2015. The results of this analysis provide valuable insight on the trend in electricity demand in the future, given that key features in China’s economic transition are likely to continue in the foreseeable future.
Among the leading factors affecting electricity demand growth, GDP continues to be the most significant driver for demand growth for power, followed by economic structural change, industrial de-leveraging, and population growth. Our estimates for the annual growth rate of electricity demand range from 2.7percent (log model) to 3.3percent (linear model), significantly smaller than the estimates published in the 13th FYP (3.6–4.8percent). The difference in electricity demand in 2020 is about 350 TWh, or roughly the generation output of 84 GW of thermal coal or 50GW of nuclear power plants using current average operating hours.
The results of this research suggest that demand growth is clearly slowing down, and there is significant uncertainty in the future growth. Such uncertainty implies great risks in investing in new generating and transmission capacity, especially under the condition of excess capacity currently existing in China’s power sector. To manage such risks, a more transparent, robust, and dynamic planning methodology and process is essential. In addition, China should consider other market instruments to help ensure reliability rather than overpaying for unnecessary capacity.
References
Acaravci, A., & Ozturk, I. (2010). Electricity consumption-growth nexus: evidence from panel data for transition countries.Energy Economics,32(3), 604-608.
Bianco, V., Manca, O., & Nardini, S. (2009). Electricity consumption forecasting in Italy using linear regression models.Energy,34(9), 1413-1421.
Ciarreta, A., & Zarraga, A. (2010). Economic growth-electricity consumption causality in 12 European countries: A dynamic panel data approach.Energy Policy,38(7), 3790-3796.
Cowan, W. N., Chang, T., Inglesi-Lotz, R., & Gupta, R. (2014). The nexus of electricity consumption, economic growth and CO 2 emissions in the BRICS countries.Energy Policy,66, 359-368.
Lin, J., He, G., & Yuan, A. (2016). Economic rebalancing and electricity demand in China.The Electricity Journal,29(3), 48-54.
Karanfil, F., & Li, Y. (2015). Electricity consumption and economic growth: exploring panel-specific differences.Energy Policy,82, 264-277.
Lin, B. (2003). Structural changes, efficiency improvement and electricity demand forecasting.Economic Research,5, 57-65. (in Chinese)
Mohamed, Z., & Bodger, P. (2005). Forecasting electricity consumption in New Zealand using economic and demographic variables.Energy,30(10), 1833-1843.
National Bureau of Statistics of the People’s Republic of China. (2017).Statistical Communiqué of the People's Republic of China on the 2016 National Economic and Social Development.
China Electricity Council (CEC). (2017)“Yearly Statistics of China Power Industry 2016”.
National Energy Agency. (2016). News Release for the 13th FYP for Electricity Development. Accessed on November 29, 2016.
Shiu, A., & Lam, P. L. (2004). Electricity consumption and economic growth in China.Energy policy,32(1), 47-54.
Xu, N., Dang, Y., & Gong, Y. (2016). Novel grey prediction model with nonlinear optimized time response method for forecasting of electricity consumption in China.Energy.
Yuan, J., Zhao, C., Yu, S., & Hu, Z. (2007). Electricity consumption and economic growth in China: cointegration and co-feature analysis.Energy Economics,29(6), 1179-1191.
Wolde-Rufael, Y. (2014). Electricity consumption and economic growth in transition countries: A revisit using bootstrap panel Granger causality analysis.Energy Economics,44, 325-330.