Globalisation, economic growth and energy consumption in the BRICS region: The importance of asymmetries

ABSTRACT

This paper examines the asymmetric impact of globalisation and economic growth on energy consumption in BRICS countries, applying the NARDL bounds approach to explore the presence of asymmetric cointegration across variables. The empirical results reveal that energy consumption is positively and negatively affected by the positive and negative globalisation shocks, respectively. A positive shock in economic growth promotes energy consumption, while a negative shock reduces energy consumption.

KEYWORDS:

• Globalisation
• growth
• energy

JEL CLASSIFICATIONS:

• F1
• O4
• Q4

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Syed Jawad Hussain Shahzad http://orcid.org/0000-0003-3511-6057

Notes

1 http://www.enerdata.net/enerdatauk/press-and-publication/enerdata-actualities/events.php.

2 There is at least one exception to the general rule where a highly politically globalised is not supporting international regulations of environment. This exception is the United States under the current Trump administration which withdrew from COP 21 recently.

3 Alam and Butt (2001) study the factor analysis for changes in energy intensity and energy consumption in Pakistan. Their results show that increases in aggregate energy intensity are due primarily to the structural effect, while increases in aggregate energy consumption are due to both the activity and the structural effect. Therefore, there may have been an inefficient use of energy in the country due to changes in its economic structure and economic activities. By applying the decomposition approach, Alam and Butt (2000) provide some insight into the changes in the economic structure and energy efficiency that occurred in Pakistan over the 1960–1998 period. They find that the cyclical component of the aggregate energy intensity index, due to the changes in economic output, decreases the efficiency of energy use by 9% per year, while the trend component, which is due to changes in consumer preferences and technology improvements, leads to increases in energy efficiency by 2.4% per annum. Alam (2002) analyses the efficiency of electricity consumption in the industrial sector in Pakistan using the multi-level decomposition model. In the case of the industrial sector, aggregate electricity intensity decreased over the 1960–1998 period, indicating improvements in the efficiency of electricity use in the industrial sector.

4 Alam and Butt (2002) investigate the causal links between energy consumption and economic growth by incorporating capital and labour as input factors in Pakistan. Their empirical analysis finds a strong long-run nexus across energy consumption, economic growth, capital, and labour. However, unidirectional causality runs from energy consumption to economic growth. Additionally, capital formation Granger causes both energy consumption and economic growth.

5 The ARDL framework can simultaneously resolve the problem of residual serial correlation and endogenous regressors if an appropriate lag order of the ARDL (p, q) model is selected (Pesaran and Shin 1998). The degree to which any endogeneity is corrected in the asymmetric ARDL framework depends on the integration order of the decomposed series and x, i.e. I(I). If the decomposed cumulative sums are I(d), then the correction is better for values of d closer to 1.

6 We do not report these results in order to conserve space, but they are available upon request from the authors.

7 We have considered both intercepts and trends when applying the ADF structural break unit root test.

8 It should be noted that in order to select the final NARDL specification, we followed the general-to-specific approach. The preferred specification, is chosen by starting with max p = 2 max q = 2 and dropping all insignificant stationary regressors. The inclusion of insignificant lags, in practice, is likely to lead to in accuracies in the estimation and may introduce noise into the dynamic multipliers.

9 Decrease or negative shock in globalisation can be assessed in following ways: Globalization decreases when countries erect barriers to trade such as tariffs, quotas etc as well as decrease in the cross-border capital flows, issuing regulations they restrict the follow of information by countries. Furthermore, globalization has had its moment and could already be in decline, steadily replaced by its successor: a new age driven by advanced robotics, artificial intelligence and additive manufacturing. These technologies stand to dramatically lower the costs of production as they become more prevalent throughout the manufacturing process.

10 The stability of NARDL estimates is tested by applying the CUSUM of square (CUSUMSQ) proposed by Brown, Durbin, and Evans (1975). The results are reported in Figure 1 and show that the plots of CUSUMSQ are between the critical bounds at the 5% significance level, indicating the reliability of the NARDL parameters.