The Choice between Renewables and Non-renewables: Evidence from Electricity Generation in 29 Countries.

• Author: Nguyen, Jeremy

1. INTRODUCTION

   In the last century, rapid global economic development has resulted in increasing demand for energy around the world. Energy consumption has risen substantially due to population expansion, urbanisation and industrialisation. For decades, electricity has been the fastest growing form of end-use energy consumption; this trend is predicted to continue for the next 20 to 30 years (International Energy Agency, 2010). World electricity generation is projected to grow 69% by 2040, from 21.6 trillion kilowatt-hours (kWh) in 2012 to 36.5 trillion kWh in 2040 (Conti et al., 2016). Reliance upon fossil fuels poses potentially irreversible adverse environmental consequences. The Paris Agreement within the United Nations Framework Convention on Climate Change (UNFCCC) has led to a commitment from 197 countries to limit the global temperature increase to 1.5[degrees]C by the end of this century (UNFCCC, 2015). To achieve the goals of The Paris Agreement, moderating aggregate energy consumption is not the only policy measure at the disposal of policy makers, changing the composition of the global energy mix and increasing the role of renewables are also critical.

   A key driver of the historical evolution of changes in energy mix relates to income levels as economies develop through time. One relevant concept in this regard is the concept of an "energy ladder" that was introduced in the literature in the late 1980s (Hosier and Dowd, 1987; Smith, 1988). Using a regression analysis of 40 developing countries from the period 1981-1984, Leach (1992) found that there is a fuel preference ladder, where households transition from using biomass fuels to modern fossil fuels such as kerosene or liquefied petroleum gas (LPG) and then to electricity. This shift occurs at a faster pace in urban areas and is strongly dependent on city size and household income. This conclusion holds even with the inclusion of renewable energy choices (micro-hydro, biogas and wind power) for households in rural areas (Barnes and Floor, 1996).

   Heltberg (2004) augmented the concept of an "energy ladder" arguing that, by moving to increasingly advanced energy options, a household may still continue to use traditional biomass fuels. In support of Leach (1992) and Barnes and Floor (1996), Heltberg (2004) highlights the strong positive relationship between modern fuel uses and per capita expenditures in households from both urban and rural areas from eight developing countries. Further, the use of modern fuels is more prevalent in urban areas and is limited to only the top-income households in rural areas. Combined, the above studies establish the stylised fact that income can bring about major changes in the relative roles that different primary sources play in providing energy. For the first time, this paper investigates such a nexus in the context of various energy sources utilized in electricity generation.

   Energy transition at country level has been examined by previous studies in the literature. Tahvonen and Salo (2001) introduced a model in which the transitions of energy mix start from primitive usage of renewables (e.g. wood), then move towards fossil-fuels, and finally to modern forms of renewables after the Industrial Revolution. For a small-scale economy, it is less costly to use primitive renewable sources of energy in comparison to non-renewables. The use of renewables then declines as economies develop and increase reliance on fossil fuels. The increase in the share of fossil-based sources typically continues until an economy reaches a higher income level enabling them to switch to modern renewable energy sources.

   Burke (2010) developed a model to explore the progress of energy mix transition, which is referred to as a ""general electricity ladder". He finds strong evidence that as a country's income increases, we usually observe a transition from domestic energy resources (hydroelectricity and oil-fired electricity generation) towards imported tradable fossil fuels (including coal, natural gas and oil), and subsequently towards electricity sourced from nuclear and renewable resources. Burke (2010) concluded that higher per capita income leads to increased demand for imported fossil fuels, while also enabling investment in modern technology such as nuclear power and wind or solar energy. In his view, the transition path is country-specific and dependent on domestic resource endowments.

   Based on the existing literature on energy transitions, income is an important determinant of the choice between different sources of energy. There exists a well-established framework for examining the relationships between general energy consumption and economic growth. In one extreme, the 'conservation hypothesis' assumes that economic growth unidirectionally causes increases in the consumption of energy. At the other extreme, the 'growth hypothesis' views the causality as running from energy consumption to economic growth. In contrast to the conservation hypothesis, in this scenario, an energy conservative policy will have negative impacts on economic growth. An alternative that allows for both, the 'feedback hypothesis' posits a bidirectional relationship between economic growth and energy consumption. Finally, the 'neutrality hypothesis' implies that there is no relationship between economic growth and energy consumption, whereby energy policy and economic growth are independent of each other (Menegaki, 2011). We can apply this framework to consider the potentially different relationships between individual primary energy sources and income, and the implications for energy mix.

   In recent years, the discussion on global energy mixes has shifted the focus towards renewable energy resources. Using panel data from 18 emerging economies, Sadorsky (2009) found that, in the long run, a 1% increase in GDP per capita leads to an approximately 3.5% increase in per capita consumption of renewables, which supports the conservation hypothesis for renewables. There are, however, studies that are in-line with the growth hypothesis for renewables. Panel data analysis for China from 1978 to 2008 showed that a 1% increase in renewable energy consumption stimulates real GDP by 0.120% and GDP per capita by 0.162%. A 1% rise in renewable energy consumption also has an effect on annual per capita income of households, with an increase of 0.444% in rural areas and a rise of 0.368% in urban areas (Fang, 2011). Similarly, a structural vector autoregressive analysis of the Indian economy in the 1960-2009 period concluded that a positive one standard-deviation shock to the consumption of renewable energy resources causes GDP to rise (Tiwari, 2011).

   The majority of recent studies find empirical support for the feedback hypothesis or a combination of different propositions for different countries with regards to renewable energy. Studies using data from OECD, Eurasian and Central American countries indicate bidirectional causality for the relationship between renewable energy consumption and economic growth (Apergis and Payne, 2010a, 2010b). Following their previous results, Apergis and Payne (2012) extended their analysis to 80 countries using a multivariate panel error correction model to include non-renewable energy. They find that there is bidirectional causality between the consumption of both renewable and non-renewable energy and economic growth, in the short and long run. Findings by Salim and Rafiq (2012) only support the feedback hypothesis in the short run. In the long run, the empirical relationship between renewable energy and growth supports the conservation hypothesis. Other studies find a mixed view regarding the dynamic of the energy-growth relationship, suggesting that outcomes vary for different countries, model specifications, energy consumption proxies and timeframes (Tugcu, Ozturk, and Aslan, 2012; Al-mulali, Fereidouni, Lee, and Sab, 2013; Bilgili and Ozturk, 2015). In the context of primary energy sources used in electricity generation, there is a consensus that renewable sources require capital-intensive technology (Borenstein, 2012; Lewis, 2007; Blanco, 2009). However, renewable technology can be competitive if the externality costs of fossil-based electricity are internalised (Weisser, 2004).

   Besides economic growth, government policy plays an important role, with the potential to either accelerate or hamper utilisation of renewable energy sources (Przychodzen and Przychodzen, 2020). Barriers to the adoption of renewable energy may come from a variety of political and institutional sources: highly controlled energy sectors, market distortions such as subsidies and tax concessions to conventional energy sources, and from uncertainty with regards to government policy (Painuly, 2001). On the other hand, policy instruments such as feed-in tariffs, tenders and tax incentives have demonstrated effectiveness for promoting the use of renewable energy in the U.S. and European Union (Kilinc-Ata, 2016). Issues of energy security also come into play; nations that are dependent on imported energy have an increased incentive to transition toward domestically produced renewable sources (Chien and Hu, 2007). Institutional barriers to increased penetration of renewable energy sources may exist with regards to financing and cost-effectiveness (Painuly, 2001). High initial capital costs and inadequate access to financing mechanisms inhibit renewable energy adoption, particularly in developing nations (Luthra, Kumar, Garg, and Haleem, 2015).

   The present study contributes to the existing literature by examining the critical role of income per capita and own- and cross-prices of primary energy sources on the use of energy specifically used in electricity production (as opposed to more general aggregate energy consumption), while distinguishing developed from developing economies. The results of this study enable us to examine a widely held conjecture...