ENERGY CONSUMPTION AND ECONOMIC GROWTH FOR SMALL ISLAND DEVELOPING STATES: A PANEL ARDL APPROACH

FAUZEL Sheereen

Senior Lecturer in Banking & Financial Services Dept of Finance & Accounting/Faculty of Law and Management University of Mauritius, Réduit, Mauritius.

Keywords:Energy consumption, Economic growth, PARDL, SIDS.

ARTICLE HISTORY: Received:19 March 2019 Revised:25 April 2019 Accepted:29 May 2019 Published:1 August 2019.

Contribution/ Originality:This study contributes to the existing literature of energy consumption-growth nexus for small island developing states. It uses new estimation methodology namely the Panel Autoregressive distributed lag model. The paper's primary contribution is finding that in both the long run and short run, there is a direct link between energy consumption and economic growth.

1. INTRODUCTION

Financial economists have been putting a lot of emphasis on the link between energy consumption, energy prices and economic growth. However, mainstream economists rather lay importance on land, labour and capital as the main factors influencing economic growth. The concept of energy use and economic growth is not thoroughly investigated for small island developing states.

SIDS countries are considered as ecologically vulnerable. These countries are small and do have certain economic disadvantages mainly in the form of limited natural resources and overdependence on a small number of goods and services for exports, high import dependence, limited local markets, and restricted ability to attain economies of scale. Also, SIDS are highly dependent on energy and hence most of their export earnings are used to finance energy expenditure.¹ Referring to CO2 emission, its average value in SIDS is 4.7 mega tons (Mt). There are 6 small island developing states which are categorized as high-income and upper-middle-income small islands which produces more than 5 Mt CO2. Trinidad and Tobago produces nearly 50 Mt of CO2 and thus represent the island which is the most polluting in the SIDS group. Together, the SIDS account for below 1 percent of global GHG emissions. However, these islands are highly affected by global climate change (SIDS Factsheet, 2013) ².

Analysing the literature, it is noted that there are mixed evidences on the link between energy consumption and economic development. For instance, Apergis and Payne (2012) investigated the case of 80 countries over the period 1990-2007, reported a two way causality between the renewable energy and economic growth and similar results are obtained for the non-renewable energy and economic growth. However, the study by Chiou-Wei et al. (2008) revealed no link between these variables. On the other hand concentrating on firms’ productivity, Aldieri and Vinci (2017) investigation reveal that there is a statistically negative influence of spatially distributed environmental spillovers on firms’ productivity  while Aldieri et al. (2018) concluded that environmental spillovers have an inverse influence by endorsing the prevalence of the displacement effect.  Studies on the impact of energy use on economic growth focus more on industrialized countries and emerging economies and less on SIDS. More so, a unified framework incorporating both energy use and carbon emission is rarely found in existing studies. In terms of techniques, no studies have been found on this topic which uses the panel autoregressive lag model.

Hence, this study aims at examining the possible linkages between energy consumption and economic growths for selected small island developing states. The causal link between CO2 emissions and economic growth is as well discussed. The methodological approach used is the Panel autoregressive distributed lag (ARDL) to cointegration as proposed by Pesaran et al. (1999).

2. ECONOMETRIC MODELING

The study aims at investigating the relationship between energy consumption and economic growth in selected Small Island Developing States³ over the period 1990 to 2016⁴ . Following the standard literature such as Levine and Zervos (1998) and Barro (1991) the following econometric model is being used;

The natural logarithm of the variables have been used in order to reduce the problem of heteroskedascity. This technique also make interpretation of the results easier and more meaningful. This result in the following:

Where i represent country, t represents time; ε is the random error term.

The parameter estimates are β1… β5 and the random disturbance term is ɛit. Data has been obtained from the World Development database. Table 1 shows the variables used in the study, with definition and sources.

Table-1. Variables definition and sources.

Variables	Definition	Related Literature	Sources
Y	GDP per capita	Han and Lee (2013); Tiwari (2011)	WDI
ENER	Primary Energy consumption	Gozgor et al. (2018)	WDI
CO2	CO2 emission per capita	Han and Lee (2013), Tiwari (2011)	WDI
HC	Human Capital	Boopen and Vinesh (2011)	UNESCO
OPEN	Trade openness	Akin (2014)	WDI
FDI	Foreign Investment	Sukar et al. (2011)	WDI

Source: Author’s Computation.

As discussed above, the Panel autoregressive distributed lag (PARDL) approach to cointegration has been used. This methodology is selected mainly based on various of its attributes. To start with, the ARDL models gives better estimates of the long run coefficients as well as unbiased estimates and valid t-statistics even when some of the regressors are endogenous (Pesaran et al., 2001; Harris and Sollis, 2003).

Thus, this paper aim at examining the link between energy consumption and economic growth in both the short-run and long-run. Hence, the MG (Mean Group) and PMG (Pooled Mean Group) estimations are used for investigation purposes. Finally, Hausman Test is being done to decide whether the MG or PMG estimators will be used.

2.1. Mean Group (MG) Estimator⁵

Pesaran and Smith (1995) first discussed about the mean group estimator. They argued on the estimation of separate regressions for individual country in the sample and to calculate the coefficients as unweight means of the estimated coefficients for each countries. By so doing, there is no imposition of any sort of restrictions. Thus all the coefficients are allowed to vary and be heterogeneous in the long-run and short-run. For the individual countries, the long run estimates are obtained from the MG framework for individual countries.

The ARDL is as following:

2.2. Pooled Mean Group (PMG) Model

Also we have the pooled mean group (PMG) estimators. Here, the short-run coefficients, the intercepts, the speed of adjustment to the long-run equilibrium as well as the error variances are allowed to be heterogeneous by country. However, the long-run slope values are limited to be homogeneous across countries.

The unrestricted specification for the ARDL system of equations for different consecutive time periods and for a number of countries for the dependent variable Y is:

All the dynamics and the ECM terms are free to vary in PMG.

2.3. Error Correction Term

The error correction term reflects the speed of adjustment which help to reestablish equilibrium in the dynamic model. This term gives an indication on how rapidly variables meet the equilibrium. The condition is that this coefficient should be negative and statically significant (Banerjee et al., 1998).

2.4. The Hausman Test

The hausman test allows us to determine whether the MG or PMG estimates should be used. In the case where the parameters are homogenous, then it imply that the the most efficient estimates are the PMG rather than the MG. Hence, the efficient estimator under the null hypothesis, is the PMG estimates. In this investigation, the hausman test shows that the PMG should be adopted as p value is greater than 5%.

Hence only the PMG results are shown in the next section and the results are discussed. Also, it was found that the some variables are stationary at level and some variables are stationary at first difference and non at second difference. Hence, the PARDL framework can be used in this study.

3. ANALYSIS AND DISCUSSION

3.1. Long Run Estimates

Analysing the long run estimates of the regression, the results shows that the energy coefficient is positive (+0.1771) and significant.  Hence, it implies that for every 1% increase in energy use, output increases by 0.1771%. It can therefore be argued that energy consumption induce economic growth in the selected small island developing states under this study. This finding is in line with the growth hypothesis which suggest that the consumption of energy consumption is very important and it influences economic growth. Also, it is a counterpart to the inputs of factors of production namely capital and labour. Hence, an increase in energy consumption causes an increase in real GDP. Therefore, it can be concluded that the economies considered in this study are ‘energy dependent’ and energy conservation policies may be applied with care as these can have adverse effects on real GDP. However, as highlighted by Costantini and Martini (2010) if the economies uses more efficient energy sources and less polluting preferences, then it can stimulate growth rather than creating an obstacle to it.

Table-2. PMG estimates (Dependent variable: LnY)

LONG RUN  RESULTS	Variable	PMG	Significance
	LnENER	0.1771	0.0001
	LnCO2	0.5553	0.0000
	LnHC	-0.1658	0.0463
	Ln OPEN	0.1358	0.0339
	Ln FDI	0.1385	0.000
SHORT RUN RESULTS	ECT	-0.1351	0.000
	D(LnENER)	0.0309	0.0334
	D(LnCO2)	-0.0055	0.8457
	D(LnHC)	0.0857	0.0712
	D(LnOPEN)	-0.0468	0.1810
	D(LnFDI)	-0.0075	0.0096
	Constant	1.1623	0.0000
	No. Countries	20
	Hausman test Prob> chi 2	0.312 P>5% - PMG Chosen

Note: D is first difference operator; PMG means pooled mean group; ECT is error correction term. Dependent variable: Economic growth (lnY) Independent variable:

Further analysing the results, it can be observed that pollutant emissions increases growth in the long run, suggesting that an increase in CO2 emissions leads to economic prosperity. Similar results were obtained for the case of Lim et al. (2014) for Philippines, where CO2 emissions were seen to boost economic growth. For instance, Munasinghe (2008) is of the view that less developed countries end up being more polluted whereby CO2 emmisions are high and this is the result of their poor knowledge and lack of technology when increasing their economic development. Munasinghe (2008) claims that green growth will be achieved by implementing the ‘sustainomics’ theory together with international collaboration. The results obtained from the present study are in line with Munasinghe (2008) since SIDS are form part of the less developed countries, enlightening the direct link between CO2 and GDP. The findings of Munasinghe (2008) also declares that if high income countries would help the low income countries through the transfer of clean technologies, funding and advices on sustainable development policies; then these least developed countries would take advantage from industrialisation and at the same time minimise environmental degradation. In contrast, Holtz-Eakin and Selden (1995) reported that in the long run, less CO2 emissions will increase GDP per capita.

Moreover, it is further detected from the results that economic growth is dependent on foreign investment. For instance an increase in foreign investment boost economic growth by 0.14%. This result is similar to Sukar et al. (2011). For instance, as summarized by Hermes and Lensink (2003) FDI bring several positive externalities to the host country through various channels. The study quoted the competition channel whereby more competition leads to more productivity, boost efficiency and investment in human and physical capital. There is also the training channel whereby FDI train labour resulting in knowledge spillover. Technological spillovers also take place as FDI come with advanced technology. There is as well the linkages channel as well as the demonstration channel where domestic firms imitate the more advanced technologies used by foreign firms. Zooming on the results, it is observed that trade openness also cause economic growth.  As argued by Silajdzic and Mehic (2018) trade openness can lead to a better resource allocation through mean of  economies of scale and scope and also through more competition. Moreover, it enables knowledge diffusion and technology transfer which results in productivity patterns that boost technological progress and result in more efficiency. In addition to the ARDL results, there is also short run estimates which are reported. Given that the variables are cointegrated provides evidence that the ECM representation can be used in order to investigate the short run dynamics. Estimation results are presented in Table 2 above. It is observed that even in the short run the energy variable is positive and significant. Hence, energy consumption do increase economic growth in both the short run and long run. Moreover, human capital is as well seen to be a determinant of economic growth in the short run.

4. CONCLUSION AND POLICY IMPLICATIONS

Using rigorous econometric panel analysis, this paper attempted to investigate the growth-environment nexus over the period 1990-2016 for selected small island developing states. Preliminary tests showed that some data are stationary at level and others are stationary at first difference and non at second difference. Consequently, the causal relationships among the variables were investigated within a PARDL framework. Also, the hausman test shows that the best estimates to use should be the pool mean group ones. The main empirical findings revealed that in the long run, there is a direct link between energy consumption and economic growth. Moreover, CO2 emissions, foreign investment and trade openness are seen to be determinants of economic growth in the long run.

These findings are very significant for designing proper environmental policies. From this investigation, it can be concluded that energy conservation policies would not benefit the economies in terms of economic growth. However, to attain sustainable development, low carbon alternatives should in used. For instance, the economies could adopt renewable energy policies. For this change to happen, collective long-term commitment of all stakeholders are required. For instance, the government along with private sectors, international agencies as well as the population should be involved to help this transmission and ensure that economic growth is not affected. An effective regulatory framework is strongly recommended. Further research should focus more on the causal links between CO2 emission and the other determinants in the study. Moreso, a static analysis can also be done to compare with the present dynamic analysis. Likewise, the research can be extended to a longer time frame. 

Funding: This study received no specific financial support.

Competing Interests: The author declares that there are no conflicts of interests regarding the publication of this paper.

REFERENCES

Akin, C.S., 2014. The impact of foreign trade, energy consumption and income on CO2 emissions. International Journal of Energy Economics and Policy, 4(3): 465-475.

Aldieri, L., V. Sena and C.P. Vinci, 2018. Domestic R&D spillovers and absorptive capacity: Some evidence for US, Europe and Japan. International Journal of Production Economics, 198: 38-49.Available at: https://doi.org/10.1016/j.ijpe.2018.01.015.

Aldieri, L. and C. Vinci, 2017. The role of technology spillovers in the process of water pollution abatement for large international firms. Sustainability, 9(5): 868.Available at: https://doi.org/10.3390/su9050868.

Apergis, N. and J.E. Payne, 2012. Renewable and non-renewable energy consumption-growth nexus: Evidence from a panel error correction model. Energy Economics, 34(3): 733-738.Available at: https://doi.org/10.1016/j.eneco.2011.04.007.

Banerjee, A., J. Dolado and R. Mestre, 1998. Error-correction mechanism tests for cointegration in a single-equation framework. Journal of Time Series Analysis, 19(3): 267-283.Available at: https://doi.org/10.1111/1467-9892.00091.

Barro, R.J., 1991. Economic growth in a cross section of countries. The Quarterly Journal of Economics, 106(2): 407-443.Available at: https://doi.org/10.2307/2937943.

Boopen, S. and S. Vinesh, 2011. On the relationship between CO2 emissions and economic growth: The Mauritian experience. In University of Mauritius, Mauritius Environment Outlook Report, 14: 2015. Available from http://www.csae.ox.ac.uk/ .

Chiou-Wei, S.Z., C.-F. Chen and Z. Zhu, 2008. Economic growth and energy consumption revisited—evidence from linear and nonlinear Granger causality. Energy Economics, 30(6): 3063-3076.Available at: https://doi.org/10.1016/j.eneco.2008.02.002.

Costantini, V. and C. Martini, 2010. The causality between energy consumption and economic growth: A multi-sectoral analysis using non-stationary cointegrated panel data. Energy Economics, 32(3): 591-603.Available at: https://doi.org/10.1016/j.eneco.2009.09.013.

Gozgor, G., C.K.M. Lau and Z. Lu, 2018. Energy consumption and economic growth: New evidence from the OECD countries. Energy, 153: 27-34.Available at: https://doi.org/10.1016/j.energy.2018.03.158.

Han, C. and H. Lee, 2013. Dependence of economic growth On CO2 emissions. Journal Of Economic Development, 38(1): 47-57.Available at: https://doi.org/10.35866/caujed.2013.38.1.003.

Harris, R. and R. Sollis, 2003. Applied time series modelling and forecasting. Wiley.

Hermes, N. and R. Lensink, 2003. Foreign direct investment, financial development and economic growth. The Journal of Development Studies, 40(1): 142-163.

Holtz-Eakin, D. and T.M. Selden, 1995. Stoking the fires? CO2 emissions and economic growth. Journal of Public Economics, 57(1): 85-101.Available at: https://doi.org/10.1016/0047-2727(94)01449-x.

Levine, R. and S. Zervos, 1998. Stock markets, banks, and economic growth. American Economic Review, 88(3): 537-558.

Lim, K.-M., S.-Y. Lim and S.-H. Yoo, 2014. Oil consumption, CO2 emission, and economic growth: Evidence from the Philippines. Sustainability, 6(2): 967-979.Available at: https://doi.org/10.3390/su6020967.

Munasinghe, M., 2008. Addressing the sustainable development and climate change challenges together: Applying the sustainomics framework. Paper Presented at the Beijing Forum, Beijing, China.

Pesaran, M.H., Y. Shin and R.J. Smith, 2001. Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics, 16(3): 289-326.Available at: https://doi.org/10.1002/jae.616.

Pesaran, M.H., Y. Shin and R.P. Smith, 1999. Pooled mean group estimation of dynamic heterogeneous panels. Journal of the American Statistical Association, 94(446): 621-634.Available at: https://doi.org/10.1080/01621459.1999.10474156.

Pesaran, M.H. and R. Smith, 1995. Estimating long-run relationships from dynamic heterogeneous panels. Journal of Econometrics, 68(1): 79-113.Available at: https://doi.org/10.1016/0304-4076(94)01644-f.

Silajdzic, S. and E. Mehic, 2018. Trade openness and economic growth: Empirical evidence from transition economies. Trade and Global Market. IntechOpen.

Sukar, A., S. Ahmed and S. Hassan, 2011. The effects of foreign direct investment on economic growth: The Case of Subsahara Africa. Southwestern Economic Review, 34: 61-74.

Tiwari, A.K., 2011. A structural VAR analysis of renewable energy consumption, real GDP and CO2 emissions: Evidence from India. Economics Bulletin, 31(2): 1793-1806.

 Appendix: Description of SIDS countries under study

Table-3. Sample of Countries.

Caribbean SIDS	Pacific SIDS	AIMS SIDS
Antigua and Barbuda Aruba Bahamas Barbados	Kiribati	Cape Verde Comoros
Belize	Papua New Guinea	Mauritius Maldives
Dominica
Dominican Republic
Grenada Guyana
Jamaica St. Lucia
St. Vincent and the Grenadines
Trinidad and Tobago

Source: Author’s Computation.

Views and opinions expressed in this article are the views and opinions of the author(s), Energy Economics Letters shall not be responsible or answerable for any loss, damage or liability etc. caused in relation to/arising out of the use of the content.

---

Footnotes: