Does Renewable Energy Convey Information to Current Account Deficit?: Evidence from OECD Countries

Abstract

Energy trade balance has been the main factor behind current account imbalances in many developed and developing countries. This study investigates whether or not renewable energy conveys information to the current account deficit of selected OECD countries. Utilizing a dataset spanning from 1990 to 2021, we apply a Panel Autoregressive Distributed Lag (ARDL) estimator to determine the interrelation of current account deficit (CAB) as a percentage of GDP with selected indicators, namely, net energy import in total final energy consumption (NEI), the share of renewable energy in total electricity production (REN_TEO), and fiscal deficit as a percentage of GDP (FAB). The results of long-term estimations reveal that as net energy import increases, the current account deficit deteriorates. On the other hand, in the case that countries utilize more of renewable energy in their total electricity generation, their current account deficits improve. Thus, we conclude that energy policy matters for the current account balances and subsequently for the well-being of OECD economies. Finally, we find strong evidence for the twin deficit hypothesis, as fiscal deficit is negatively interrelated with current account deficit both in the short-run and long run. In other words, an increase in the level of budget deficit is associated with an upsurge in the current account deficit problem. Furthermore, the Dumitrescu-Hurlin causality test reveals that there is bidirectional heterogeneous causality between current account deficit and budget deficit. Additionally, when the countries in the sample are grouped by their per capita GDP levels, estimations reveal that the direction of interaction between CAB and energy-related indicators (NEI and REN_TEO) does not differ between Group 2 (the ones whose per capita incomes are over USD 25,000 but below USD 50,000) and Group 3 (the ones having more than USD 50,000 per capita income) countries. However, the coefficients of energy-related indicators for Group 2 countries are higher than those of Group 3 ones, suggesting that energy policy matters more for Group 2 countries’ current account imbalances in the long-term.

1. Introduction

Current mitigation pledges fall short of achieving global temperature goals by 2030. The implementation gaps also persist. This shows the need for comprehensive and concrete policies as well as coordinated efforts to implement them. However, the challenge lies in adjusting the pace of climate policies without jeopardizing the growth prospects, specifically in the case of developing economies. At the current juncture, with dramatically increasing energy prices, energy security concerns are dominating efforts to decarbonize in the near to medium term.

The IMF, in its report of [1], points out that the combined cost of the energy crisis and the cost of reaching the Nationally Determined Contributions (NDCs) is higher than the cost of reaching the NDCs alone in many countries, except in oil-exporting ones. This increase in costs is mainly due to the loss of access to the relatively cheap natural gas and its substitution by more expensive imports. Thus, for energy security purposes, transitioning away from fossil fuels would not be feasible in the near to medium term.

However, in the long term, low-carbon transition does not substitute for energy security but rather helps sustain energy security. A safe, environmentally sound, and economically viable energy pathway that will sustain human progress into the distant future is clearly imperative and also possible, as announced way back earlier in the United Nations report of [2]. Furthermore, United Nations Sustainable Development Goals 7 (SDG-7) focus on ensuring access to affordable, reliable, sustainable, and modern energy for all. In order to reach this goal, Target 7.2 aims to substantially increase the share of renewable energy in the global energy mix by 2030.

Checking the recent global outlook, it is recognized that post-COVID, while the predictability in commodity markets has decreased, global geopolitical uncertainties have substantially increased. Then, with the Ukraine-Russia war, energy import dependency has become an important concern around the globe, leading many countries to increase the share of domestic and renewable energy resources such as wind, solar, hydraulic, geothermal, biomass, and biogas in their energy portfolios as well as in electricity generation. These efforts also play a major role in the global decarbonization process towards achieving net zero emissions by 2050. Furthermore, thanks to recent technological advances and significant decreases in renewable energy installation and operating costs, jurisdictions’ interest in renewable energy has been growing up.

According to the IRENA figures depicted in [3], renewable energy capacity around the world more than doubled in the 10 years between 2013 and 2022. While global renewable capacity was 1566 gigawatts in 2013, it reached 3371 gigawatts in 2022. However, progress has been quite uneven across the world, with large capacities built up in Europe, as depicted by a 69% growth rate in the continent during the stated period. There is also huge capacity progress in China. China’s renewable energy capacity in 2022 is almost 3.5 times as big as its capacity in 2013. Figures indicate that global renewable capacity surges are largely driven by Europe and China.

In this regard, the recent energy crisis stressed the importance of diversifying the energy mix, in favor of renewable energy. Therefore, based on the IEA report of [4], current elevated energy prices have boosted investment in clean energy, with 14.8% growth in 2021–2022 and 7.6% in 2022–2023 periods, driven by renewables and electric vehicles. Although renewable energy has reached unprecedented global growth rates over the last decade, its share in total final energy consumption has remained steady—at around 17 percent. This was mainly because of the fact that global energy consumption grew at a similar rate. Thus, it will definitely take time to accumulate adequate low-carbon energy capacity and increase the share of renewable energy in total final consumption. Furthermore, global energy statistics [5] indicate that while fossil-based sources dominate all other sources with 61.4% of total electricity output, 28.5% of the world’s total electricity is generated using renewable sources.

On a positive note, as a response to the energy crisis, in addition to increasing alternative energy sources, countries also adopt measures to reduce energy demand via improving efficiency. In an attempt to cut energy demand, countries raise national energy efficiency targets and develop policies to serve for this purpose through some fiscal and monetary incentives. More recently, at the United Nations (UN) 28th Conference of Parties (COP28) in Dubai, 118 countries have signed the global renewables and energy efficiency pledge as a testimony of their dedication to climate action.

It is undoubted that renewable energy produced domestically and being inexhaustible will substantially improve energy independence by reducing the share of imported energy. Thus, renewable energy may also be related to a country’s economy, and studies are carried out to investigate this relationship.

In this study, our main research question focuses on determining whether or not renewable energy-related indicators convey information to the level of current account deficit in selected OECD countries. We also test if the twin deficit hypothesis holds for the sample. More specifically, we investigate the interaction between current account deficit as a percentage of GDP and the share of net energy imports in total final energy consumption, the share of renewable energy in total electricity production, as well as fiscal deficit as a percentage of GDP. We decided to investigate the current account balance as the dependent variable, as it carries extremely valuable information about the state of the economy under consideration and is closely monitored by policymakers. In addition, it is known that energy trade balance has been the main factor behind the current account deficits of many developed and developing countries. Thus, we use energy-related indicators as independent variables to explain the current account deficit. Finally, we also aim to test the twin hypothesis, using fiscal deficit as one of the explanatory variables.

The research contributes to the existing body of literature in three dimensions. First, the study differs from other studies as it investigates the relationship among selected variables, specifically for OECD economies, and by additionally grouping sample countries based on per capita GDP levels. Earlier studies investigated the relationship of renewable energy consumption with some other macroeconomic variables, such as economic growth, while some scholars studied the interaction of renewable energy consumption with the current account deficit. Contrary to earlier studies that investigate the relationship of renewable energy-related indicators with selected macroeconomic indicators, our study also tests the twin deficit hypothesis, utilizing country-based annual data from 1990 to 2021. Last but not least, the study establishes the link between energy policies and climate change, as switching to clean sources of energy will definitely help address this problem.

The remainder of this study is ordered as follows: Section 2 reviews the literature and discusses current account balance and its main sources. Section 3 examines the selected indicators of sample countries in detail. The data are presented in Section 4, while the model and methodology are discussed in Section 5. Section 6 discusses estimation results. Finally, Section 7 concludes, proposing policy recommendations.

2. Literature Review

The balance of payments of a country is comprised of 4 account types: current account, capital account, financial account, and the net errors and emissions account. These accounts reveal detailed information about countries’ foreign economic activities and financial relations. While the current account shows the difference between imports and exports of a country’s goods and services, the capital account tracks capital flows between countries. The financial account keeps track of foreign investments, loans, and other financial transactions. Net errors and omissions accounts are used when the accounts do not match and adjust the balance of payments to achieve accounting balance.

Current account imbalances have been an important concern for developing economies since the global financial liberalization of the 1980s. However, specifically after the Global Financial Crisis of 2008, current account deficits also became a problem for advanced economies. If the level of current account deficit exceeds the critical threshold of 5–6% of GDP for recurring years, it might signal a potential future crisis, specifically an exchange rate crisis. Furthermore, as the current account deficit needs to be financed by foreign direct investments, portfolio investments, and other investments, this might lead countries to be dependent on foreign resources.

In the literature, studies on current account balance are divided into 2 groups. While one group focuses on testing the twin deficit hypothesis, the other group tries to find the drivers of the current account balance.

In more detail, the ones in the first group of the literature focus on the twin deficit hypothesis that suggests current account deficit and budget deficit move together. These studies test the twin deficit hypothesis and try to capture the relationship between current account deficit and budget deficit. There are 3 theoretical arguments for the twin deficit hypothesis. While the traditional Keynesian view argues for a causality running from the budget deficit to the current account deficit, the Ricardian Equivalence Hypothesis claims that there is no relationship between the two. Finally, the Current Account Targeting by [6] posits the idea that causality is from current account deficit to budget deficit, contrary to Keynesian view.

On the other hand, a second group of studies tries to find the drivers (determinants) of current account balances, using different variables and applying various methodologies. In these studies, lagged current account deficit, economic growth, budget deficit, real interest rate, real exchange rate, investments, savings, and financial openness are among the indicators investigated. As an example of these studies, the authors of [7] focus on OECD economies and try to find the determinants of current account deficits. Applying a Panel VAR approach to 28 OECD economies in the period of 1999–2009, they find that growth, interest rate, and budget deficit have a small and medium-term effect on the current account deficit, while the exchange rate does not have any effect for these economies.

Since our main investigation point in this study is to understand the relationship between energy and current account balance, rather than determining the drivers of current account deficit, we focused more on the line of literature that discusses the link between current account deficit and various energy-related indicators.

Domestic and international literature that we searched through includes, but is not limited to, 5 main headings: i. Renewable Energy Consumption-Economic Growth, ii. Renewable Energy Consumption-Economic Growth-Current Account Balance, iii. Renewable Energy Consumption-Energy Import-Current Account Balance, iv. Renewable Energy Consumption-Economic Growth-Environmental Degradation, and v. Renewable Energy Consumption-Current Account Balance-Environmental Degradation

A summary of the literature reveals that most of the studies use annual data and apply the Autoregressive Distributed Lag (ARDL) bounds test to capture the interrelation between variables. However, country-based studies utilizing Panel ARDL or Augmented Mean Group (AMG) estimators also exist in the literature.

Many studies in the domestic literature analyze the relationship between economic growth and energy consumption. For instance, the author of [8] analyzes the relationship between economic growth and renewable energy consumption in Türkiye. Utilizing annual data covering 1960–2017 periods, the study investigates the relationship between renewable energy consumption, financial development, and economic growth. Based on the Non-linear Autoregressive Distributed Lag (NARDL) long-term estimation results, both the increases in renewable energy consumption and in financial development are associated with an increase in economic growth.

The authors of [9] used the Vector Error Correction Model (VECM) to analyze the interrelation between renewable energy consumption and economic growth in Türkiye, in the period of 1990–2019. They find a positive Granger causality running from renewable energy to economic growth. They argue that increasing the share of renewable energy investments would support Türkiye’s sustainable development.

For our specific research interest, we mostly focused on the literature that investigates the interrelation of current account deficit with some other variables. In this regard, the authors of [10] add the current account deficit to the econometric analysis, using the data spanning from 1975 to 2009. Based on the Johansen co-integration test, they find a long-term relationship between economic growth and energy consumption, as well as between economic growth and the current account deficit. While the direction of the causality is from energy consumption to growth and quite strong, there is a bidirectional but weak relationship between growth and current account deficit.

Other studies in the literature also consider the fact that energy imports are quite important in understanding the dynamics of current account balance. Thus, they add energy imports to the econometric analysis. For instance, the authors of [11,12,13,14] studies focus on the interrelation between energy imports and current account balance.

In more detail, the author of [11] investigates the impact of renewable energy consumption and energy imports on the current account balance. The study selects 11 countries, which are the highest energy importers in the period of 1995–2015. Based on the cross-sectional dependency and slope homogeneity tests, a Panel Augmented Mean Group (AMG) estimator is employed to find the long-term slope coefficients. According to the estimation results, the effect of renewable energy consumption on the current account balance is positive. It has also been found that energy imports deteriorate the current account imbalances. Dumitrescu-Hurlin (DH) causality tests show that there is a bidirectional relationship between renewable energy consumption and current account balance. Similarly, there is bidirectional causality between energy import and current account balance.

The authors of [13] also analyzed the impact of renewable energy on energy import dependency in Türkiye. Utilizing 1990–2018 annual data and applying the ARDL bounds test, they find that while renewable energy-based electricity production, GDP per capita, urban population growth, and world natural gas prices have strong interrelation with energy import dependency, world oil prices do not. They infer that renewable energy production reduces energy import dependency, while the most important determinant of this indicator is the GDP per capita.

The authors of [14] rather focus on investments and investigate the impact of renewable energy investments on energy import dependency in Türkiye. Based on ARDL estimation results, there exists a long-term relationship between the variables. They find that a 1% increase in renewable energy investments will be leading to a 0.0041% decrease in energy imports.

The author of [15] investigates the relationship between renewable energy production and import demand, utilizing panel data. The study concludes that renewable energy production helps decrease import dependency and supports debt and economic sustainability. The author argues that renewable energy will also improve energy supply security, as it will decrease fossil-based imports.

In another interesting study, the authors of [16] utilize a scenario approach and find that till 2025, almost USD 17 billion in energy-related savings will be achieved in Illinois, US, in the case that 100% renewable energy was used. They are in favor of renewable energy, as it reduces energy import dependency.

In their study, the authors of [17] focus on understanding the factors that affect the energy import demand in European countries. Utilizing an annual dataset of 1990–2015, they apply a NARDL methodology and find that the GDP growth variable has the most important implication on energy imports. They also infer that a reduction in fossil fuel consumption is associated with a decrease in energy imports. Thus, they argue for the positive role of renewable energy for the sustainable development of European countries.

The authors of [12] also utilized country-based data covering the 1990–2018 period and employed the Panel Fixed Effects model. They find that an increase in the renewable energy share in more urbanized countries will cause a higher level of decrease in energy import dependency compared to less-urbanized ones. They also make a scenario analysis, specifically for Türkiye. They argue that positive developments in renewable energy share and energy efficiency will lead to a USD 21 billion improvement in Türkiye’s current account balance in 2030.

In their unique study that utilizes monthly data spanning from 2016 to 2022, the authors of [18] investigate the interrelation among renewable energy production, installed power, current account balance, and net energy imports. In the model that defines current account balance as the dependent variable, they find co-integration among the variables. They employ a Toda-Yamamoto causality test and find that there is one-way causality between renewable energy production and current account balance. There is also one-way causality between installed power and current account balance. The long-term interrelation between current account balance and net energy imports is found to be negative. They infer that energy policies that favor renewable energy would support Türkiye to decrease its current account deficit and improve its economic sustainability.

The authors of [19] also argue that renewable energy would improve Türkiye’s current account balance. They employ a VAR methodology to investigate the relationship between renewable energy consumption and current account deficit in Türkiye for the period of 1980–2012.

The authors of [20] investigate the interrelation of renewable energy consumption and current account deficit in some European countries in the period of 1995–2018. They utilize panel data and find that an increase in renewable energy consumption is associated with a decrease in current account deficit and energy import dependency in many countries.

The author of [21] examines the so-called vulnerable 5 countries (Brazil, India, Türkiye, South Africa, and Indonesia) in the period of 2000–2017, utilizing panel data. The study investigates the interrelation between renewable energy consumption and current account deficit. While he finds a positive long-term relationship between two indicators, economic growth, the external trade deficit, oil prices, and the real exchange rate also have an impact on the current account deficit. Similarly, the authors of [22] find that renewable energy usage improves the current account deficit in OECD countries in the long-term.

In another study, the authors of [23] examine the relationship between renewable energy consumption and current account balance in G20 countries in the period of 1976–2019. They investigated the dynamic and causal relationship among renewable energy production, current account balance, energy imports, renewable energy consumption, and economic growth through the panel fourier bootstrapping ARDL model. They find that there is a unidirectional causality from energy imports, renewable energy consumption, and current account deficit to renewable energy production. They argue that if countries decrease their dependence on energy imports, they might increase the quality of the environment through the production of more renewable energy and also reduce current account imbalances.

Going forward, as climate change has emerged as an important global problem, many studies started to integrate variables into their models to proxy environmental degradation. Ref. [24] is one of these studies. The study focuses on industrialized countries and investigates short-term causality among renewable energy production, economic growth, and carbon dioxide (CO₂) emissions. Based on Granger causality tests, an increase in clean energy production will lead to a reduction in CO₂ emissions in Chile, Australia, and Austria, but an increase in Denmark and Holland’s CO₂ emissions in the long-term.

The authors of [25] also try to understand the interrelation among renewable energy, environmental degradation, and economic growth in OECD member countries. They find that renewable energy has a positive impact on economic growth and environmental degradation up to a certain threshold, but after that threshold, the impact is negative.

Finally, the authors of [26] utilize dynamic ARDL to investigate the relationship between current account balance, renewable energy consumption, and CO₂ emissions in the period of 1981–2020. The study is a country-based one, using US, China, and India data. They find that increased use of renewable energy and improvements in current account balance help decrease CO₂ emissions.

3. An Overview of Selected Indicators for the Sample

Current Account Balance, Energy Trade Balance, and Comparison of Countries Based on Selected Indicators

Before describing the data utilized in the study, this section aims to establish the link between energy trade balance and current account balance.

Current account imbalances of many countries often result from the large deficits in the goods trade balance, which is categorized under 3 parts: energy, unprocessed gold, and core goods. A deficit occurs in the goods trade balance if imports of goods are higher than exports of goods, namely if the country is a net importer. Moreover, as energy expenditures constitute a big portion of the goods trade balance, countries with high current account deficits are the ones whose energy trade balances are sizably negative. Considering this link between energy trade balance and current account balances, it might be noted that energy policies have significant importance for the economies. Countries whose energy import dependencies have historically been high are quite sensitive to global energy prices as well as to the changes in exchange rates. These effects are elevated for more vulnerable countries that have unstable macroeconomic indicators.

Checking the data of selected OECD countries in our sample, it is recognized that when net energy imports increase, current account imbalances deteriorate. For instance, historical data of [27] depict that Türkiye would have given a current account surplus in most months if energy imports were not existent.

Digging the data of selected OECD economies in our sample, it is seen that the net energy imports of Chile, Türkiye, and Greece are in the range of 55–64% of their total final energy consumption and surpass the rest of the sample. US and New Zealand are in the lowest range in terms of average net energy import as a percentage of total final energy consumption throughout the sample period. While New Zealand is utilizing renewable energy sources to a greater extent with an 81% share in its total electricity output, the US generates 61% of its total electricity using fossil-based sources, 19% and 18%, utilizing nuclear power and renewable sources, respectively. We understand from these figures that exchange rates also play a major role in net energy import levels.

Furthermore, it is seen that Greece, Türkiye, and Chile are not adequately utilizing their renewable sources in electricity generation, while New Zealand and Costa Rica’s share of renewable sources in total electricity output is 99% and 81%, respectively. We might argue that the US, Türkiye, Poland, Greece, and Estonia might improve their utilization of renewable sources, as fossil-based electricity production in these countries is above the world average of 61.9%. Although some countries (Czechia, US) also started to utilize nuclear energy to cut their energy imports and expenses, fossil-based electricity production still dominates the picture.

Checking the fiscal deficit data of the sample, it is seen that the fiscal deficit as a percentage of GDP of almost all countries is well above the average level of the country groups they belong to: advanced, emerging, and middle-income. In 2022, the average fiscal deficit as a percentage of GDP of advanced and emerging and middle-income countries is 3.15% and 4.89%, respectively. In the post-COVID period and specifically after the global energy crisis experienced due to the Russia-Ukraine conflict, additional expenses led to narrow fiscal spaces in both advanced and emerging and middle-income economies.

4. Data

In an attempt to make a robust analysis, the countries included in our sample are selected based on a determined logic. Out of 38 OECD member countries, 15 were selected based on satisfaction of 3 criteria. Countries that have i. positive net energy import, ii. negative current account balance, and iii. negative fiscal balance were included in our sample. Chile, Costa Rica, Czechia, Estonia, Greece, Hungary, Latvia, Lithuania, New Zealand, Poland, Portugal, Slovak Republic, Türkiye, UK, and US were the countries that satisfied the above 3 criteria simultaneously and were included in our sample.

In our second set of panel regressions, we classified these 15 countries into 3 groups, based on their per capita GDP levels. 1st group includes countries whose GDP per capita is lower than USD 25,000. The 2nd group includes countries with a per capita income level over USD 25,000 but below USD 50,000. Finally, the 3rd group includes countries that have more than USD 50,000 per capita incomes.

Since the national energy balances data are only available in annual frequency, our sample includes country-based annual data of 15 OECD countries in the period of 1990–2021. The time series data of each country’s net energy import, total final energy consumption, as well as electricity generation by source (fossil-based, renewable-based, and nuclear-based) indicators are available in IEA. Fossil-based sources include coal, oil, and natural gas, while renewable sources consist of hydro, geothermal, solar PV, solar thermal, wind, tide, and biofuels.

Table 1. Data Description and Sources.

Indicator	Definition	Unit	Data Source
Current Account Balance	CAB	% of GDP	[28]
Net Energy Import	NEI	% of total final energy consumption	[29]
Share of Renewable Sources	REN_TEO	% of total electricity output	[30]
Fiscal Balance	FAB	% of GDP	[31]

summarizes the description and sources of data utilized in the study.

The details of the variables included in the model are as follows:

• Current Account Balance, as a % of GDP;

Current account balance is the sum of net exports of goods and services, net primary income, and net secondary income. The data are sourced from the International Monetary Fund, Balance of Payments Statistics Yearbook and data files, and World Bank and OECD GDP estimates. We accessed the data through the World Bank World Development Indicators dataset.

• Net Energy Import, as a % of Total Final Energy Consumption;

The data are sourced from the IEA Energy Statistics Data Browser. Net energy imports are estimated as energy use less production, both measured in oil equivalents. A negative value indicates that the country is a net energy exporter. Energy use refers to the use of primary energy before transformation to other end-use fuels, which is equal to indigenous production plus imports and stock changes, minus exports and fuels supplied to ships and aircraft engaged in international transport. Total final energy consumption figures are also sourced from the IEA Energy Statistics Data Browser. We calculate net energy import as a percentage of total final energy consumption.

• Share of Renewable Sources in Total Electricity Output;

According to [32], renewable energy receives widespread and strong support from the public as it contributes to reducing energy import dependency, helps protect the environment via reducing CO₂ emissions, and increases employment as they are domestic resources. The International Energy Agency (IEA) considers that renewables are at the center of the transition to less carbon-intensive and more sustainable energy systems. The source of data utilized in this study is IEA, World Energy Statistics and Balances, 2023. We calculated the share of renewables in total electricity output by dividing the renewable-sourced electricity generation by total electricity generation by all sources. Renewable-based electricity production includes the generation of electricity using hydro, solar PV, solar thermal, geothermal, wind, tide, and biomass energy as the primary sources of energy.

• Fiscal Balance as a % of GDP;

The overall fiscal balance refers to net lending (+)/borrowing (−) of the general government. When the balance is negative, the government has a fiscal deficit. When the balance is positive, the government has a fiscal surplus. Data are sourced from IMF Fiscal Monitor, 2024.

Descriptive statistics of the variables are presented in

Table 2. Descriptive Statistics.

	CAB	NEI	REN_TEO	FAB
Mean	−0.03	0.39	0.30	−0.03
Min	−0.21	−0.18	0	−0.15
Max	0.08	0.69	0.99	0.08
Std. Dev.	0.04	0.20	0.29	0.04
Skewness	−0.78	−0.51	0.87	−0.27
Kurtosis	4.86	2.49	2.51	3.54
Observations	447	480	480	421

, as follows:

Despite the fact that our study does not focus on any kind of cause and effect relationship between the variables, the correlation among the variables is also checked. As it is seen in

Table 3. Correlation Matrix.

	CAB	NEI	REN_TEO	FAB
CAB	1
NEI	−0.18	1
REN_TEO	0.09	0.14	1
FAB	0.06	−0.15	0.22	1

, variables included in the model have quite low levels of correlation.

5. Model and Methodology

This study’s empirical approach is based on the following model and applied to panel data of 15 OECD countries. Model variables are used in differenced forms based on the results of unit root tests.

$${\mathrm{CAB}}_{it}={\mathrm{a}}_0+{\mathrm{a}}_1{\mathrm{NEI}}_{it}+{\mathrm{a}}_2\mathrm{REN}\_{\mathrm{TEO}}_{it}+{\mathrm{a}}_3{\mathrm{FAB}}_{it}+e_{it}$$

The expected sign for a₁ is negative, as a percentage increase in net energy imports is expected to deteriorate the current account deficit. We expect to test the Keynesian view on the twin deficit hypothesis, thus our ex-ante expectation for a₃, the coefficient of fiscal deficit is also negative. On the other hand, our main inquiry in this study is to investigate the interrelation of renewable energy with current account balance. Our ex-ante expectation is that if the share of renewable-based sources in total electricity production (REN_TEO) increases, the current account deficit will improve. Since the sign of our dependent variable is in negative terms, we expect coefficient a₂ to be positive.

Due to the existence of cross-sectional dependence, slope heterogeneity, and non-stationarity in the dataset, a Panel Autoregressive Distributed Lag (ARDL) model, namely the Pooled Mean Group (PMG) estimator, is employed. The estimator was proposed by the authors of [33] and supposes homogeneous long-run equilibrium across countries and heterogeneous short-run relationships. Differences among countries explain the heterogeneous short-run behavior.

The PMG model is particularly suited for dynamic panels with non-stationary variables, allowing to model both short-term and long-term relationships while accounting for heterogeneity across countries in the short-run. PMG estimator allows for the short-run dynamics and error variances to differ among countries while assuming homogeneity in long-term relationships. Since the study covers OECD countries, which have diverse economic conditions and energy policies, PMG not only allows us to explore long-term trends but also helps recognize short-term country-specific adjustments.

The PMG model finds short-run coefficients $\lambda , {\lambda }^{\prime }, {\lambda }^{″}, {\lambda }^{‴}$ that are associated with lagged dependent variables and explanatory variables. The model also reveals long-run coefficients ($\theta$ type coefficients) for explanatory variables. The speed of adjustment, namely the error correction term ${\varphi }_{1i}$ is also proposed by the model to show the pace of adjustment between short-term and long-term behavior. The authors of [33] state that the pace of adjustment, namely the error correction term, should be negative to indicate that a disequilibrium is adjusted in the long-run. According to [34], this approach is more consistent in generating long-run coefficients regardless of whether the order of integration is I (0) or I (1). In their study, the authors of [35] also argue that the use of Panel ARDL is appropriate as it is more efficient and consistent with the existence of long-run relationships.

In this study, the Panel ARDL Model is represented as:

$$CABit=a_{1i}+{\sum }_{l=1}^p{a}_{10}{CAB}_{it-l}+{\sum }_{l=0}^q{a}_{11}{NEI}_{it-l}+{\sum }_{l=0}^q{a}_{12}{REN\_TEO}_{it-l}+{\sum }_{l=0}^q{a}_{13}{FAB}_{it-l}+{e1}_{it}$$

After parameterization, the previous equations are represented as follows:

$$\begin{array}{l}\Delta {CAB}_{it}=a_{1i}+{\varphi }_{1i}({CAB}_{it-l}+{\theta }_{11}{NEI}_{it-l}+{\theta }_{12}{REN\_TEO}_{it-l}{+\theta }_{13}{FAB}_{it-l})\\ +{\sum }_{l=1}^{p-1}{\lambda }_{1il}{\Delta CAB}_{it-l}+{\sum }_{l=0}^{q-1}{{\lambda }^{\mathrm{\prime }}}_{1il}{\Delta NEI}_{it-l}+\sum _{l=0}^{q-1}{\lambda ″}_{1il}\Delta {REN\_TEO}_{it-l}+\sum _{l=0}^{q-1}{\lambda ‴}_{1il}{\Delta FAB}_{it-l}+{e1}_{it}\end{array}$$

The list of symbols used in equations is available in Appendix A,

Table A1. The List of Symbols.

Symbol	Description
CABit	Current Account Balance (as a % of GDP) of Country i at time t
NEIit	Net Energy Import of Country i at time t
REN_TEOit	Share of Renewable Sources in Total Electricity Output of Country i at time t
FABit	Fiscal Balance (as a % of GDP) of Country i at time t
$\Delta {\mathrm{C}\mathrm{A}\mathrm{B}}_{\mathrm{i}\mathrm{t}}$	Differenced Form of CABit
$\Delta {\mathrm{N}\mathrm{E}\mathrm{I}}_{\mathrm{i}\mathrm{t}}$	Differenced Form of NEIit
$\Delta {\mathrm{R}\mathrm{E}\mathrm{N}\_\mathrm{T}\mathrm{E}\mathrm{O}}_{\mathrm{i}\mathrm{t}}$	Differenced Form of REN_TEOit
$\Delta {\mathrm{F}\mathrm{A}\mathrm{B}}_{\mathrm{i}\mathrm{t}}$	Differenced Form of FABit
$\mathsf{\theta } \mathrm{t}\mathrm{y}\mathrm{p}\mathrm{e} \mathrm{c}\mathrm{o}\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{s}$	Long-run coefficients
$\mathsf{\lambda } \mathrm{t}\mathrm{y}\mathrm{p}\mathrm{e} \mathrm{c}\mathrm{o}\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{s}$	Short-run coefficients
ϕ	Error Correction Term
Subscript l	Lag value of the variable

.

6. Estimation Results and Discussion

6.1. Cross-Sectional Dependence Test

Cross-sectional dependence in the data was tested with Breusch-Pagan LM, Peseran Scaled LM, and Peseran CD tests. The test results are shown in

Table 4. Cross-sectional Dependence Test.

	Breusch-Pagan LM	Peseran Scaled LM	Peseran CD
Variable
CAB	(672.54) [0.00] *	(39.16) [0.00] *	(15.03) [0.00] *
NEI	(975.00) [0.00] *	60.03 [0.00] *	(−0.25) [0.79]
REN_TEO	(1365.00) [0.00] *	(86.95) [0.00] *	(22.13) [0.00] *
FAB	(505.88) [0.00] *	(27.66) [0.00] *	(16.60) [0.00] *
Residuals	(649.98) [0.00] *	(37.61) [0.00] *	(14.89) [0.00] *

Note: H0: No cross-sectional dependence (correlation) (.) and [.] indicate test statistics and probability values, respectively. * represents the level of significance at 5%.

. Since the probability value shows a 5% level of significance in all individual variables, the null hypothesis of no cross-sectional dependence is rejected based on at least 2 tests. Thus, we report that there is cross-sectional dependence in all series. Furthermore, after estimating a simple ordinary panel regression, we also confirm that residuals are also cross-sectionally correlated.

6.2. Slope Homogeneity Test

The slope homogeneity test of [36], called the Delta Test, and its adjusted version reveal that slope coefficients of all individual variables as well as the overall model are heterogeneous. Slope homogeneity test results are depicted in

Table 5. Slope Homogeneity Test.

Delta	Overall Model	NEI	REN_TEO	FAB
∆ˆ	(10.89) [0.00] *	(4.74) [0.00] *	(8.97) [0.00] *	(7.47) [0.02] *
∆ˆ adj.	(12.04) [0.00] *	(5.05) [0.00] *	(9.55) [0.00] *	(7.96) [0.01] *

Note: H0: slope coefficients are homogenous. (.) and [.] indicate test statistics and probability values, respectively. * represents the level of significance at 5%.

.

6.3. Unit Root Test

Under the existence of cross-sectional dependence, we apply the cross-sectionally augmented Im, Pesaran, and Shin (CIPS) test of [37], to test for the stationarity in panel models. CIPS test statistics are the sample averages of the individual cross-sectionally augmented ADF (CADF) statistics. The results of the CIPS tests for the panel are presented in

Table 6. CIPS Unit Root Test.

	Level	First Difference
Variable
Constant Option
CAB	(−2.10)	(−3.94 )*
NEI	(−2.06)	(−5.03) *
REN_TEO	(−1.76)	(−4.89) *
FAB	(−1.96)	(−3.56) *
Constant and Trend Option
CAB	(−2.29)	(−3.83) *
NEI	(−2.94) *
REN_TEO	(−3.25) *
FAB	(−2.47)	(−3.71) *

Note: If the test statistics are more positive or more negative than the critical values, we reject the hypothesis. Ho: There is a unit root. * indicate level of significance at 5%. The critical value at 5% significance is −2.26 for the constant option and −2.78 for the constant and trend option.

. Both constant as well as constant and trend options of the CIPS test results indicate that CAB and FAB variables are stationary in the first difference. However, it appears that two options of the CIPS test propose different stationarity levels for REN_TEO and NEI variables. Visualizing the charts of these 2 variables as well as testing whether or not trend exists, we find that there was not a clear sign of trend for these variables. To be on the safe side, while running the model, we used first differenced versions of all variables.

6.4. Cointegration Tests

While the test of [38] and the tests of [39,40,41] cointegration tests reveal cointegration in all panels, the test of [42] finds that some panels are cointegrated. However, overall, the hypothesis of no cointegration is rejected. The results of cointegration tests are shown in

Table 7. Cointegration Tests.

Ho: No Cointegration	Statistic	p-Value
[38]—Kao Test- Ha: All panels are cointegrated
Modified Dickey-Fuller t	−3.20	0.0007 *
Dickey-Fuller t	−3.27	0.0005 *
Augmented Dickey-Fuller t	−3.70	0.0001 *
Unadjusted modified Dickey Fuller t	−6.74	0.0000 *
Unadjusted Dickey-Fuller t	−4.63	0.0000 *
[39,40,41]—Pedroni Test- Ha: All panels are cointegrated
Modified Phillips-Perron t	−3.18	0.0007 *
Phillips-Perron t	−5.68	0.0000 *
Augmented Dickey-Fuller t	−4.39	0.0000 *
[42]—Westerlund Test
Variance Ratio:
Ha:Some panels are cointegrated	−2.32	0.010 *
Ha:All panel are cointegrated	−1.21	0.110 *

* indicate level of significance at 5%.

.

6.5. PMG Estimation Results

To estimate the coefficients in our model, we apply the pooled mean group estimator of [33], namely the panel ARDL estimator. The PMG estimation results are reported in

Table 8. PMG Estimation Results.

CAB		All Sample 1	Group 1 2	Group 2 3	Group 3 4
	Variable
Long-run relationship	NEI	(−0.105) [0.000] *		(−0.313) [0.000] *	(−0.099) [0.000] *
	REN_TEO	(0.048) [0.001] *		(0.166) [0.000] *	(0.037) [0.013] *
	FAB	(−0.312) [0.000] *		(−0.104) [0.247]	(−0.296) [0.000] *
Error Correction term		(−0.390) [0.000] *		(−0.357) [0.000] *	(−0.439) [0.317]
Short-Run Relationship	d_NEI	(−0.021) [0.667]		(−0.014) [0.795]	(−0.057) [0.417]
	d_REN_TEO	(0.078) [0.215]		(0.113) [0.149]	(−0.017) [0.885]
	d_FAB	(−0.319) [0.007] *		(−0.347) [0.014] *	(−0.060) [0.000] *
	Constant	(−0.0098) [0.015]		(0.016) [0.164]	(−0.014) [0.278]
Residuals		I(0)		I(0)	I(0)

Note: H0: Coefficient is equal to 0, (.) and [.] indicate coefficient and probability values, respectively. * represents the level of significance at 5%. ¹ All sample includes 15 OECD countries and is the main estimation in the study. ² Group 1 includes countries whose GDP per capita is lower than USD 25,000. Since there is only one country, Costa Rica, in this group, the estimations are not reported. ³ Group 2 includes countries with a per capita income level over USD 25,000 but below USD 50,000. ⁴ Group 3 includes countries that have more than USD 50,000 per capita incomes.

for all 15 countries in the sample as well as for the 3 classifications of country groups, as specified in Section 4.

In line with our ex-ante expectations, the results of long-term estimations including all 15 countries in the sample reveal that as net energy import increases, current account deficit deteriorates. As energy trade balance is the main source of current account deficits in many OECD countries, this finding is not surprising.

On the other hand, in the case that countries utilize more of renewable energy in their total electricity output, their current account deficit levels improve, as the positive sign of the coefficient of the REN_TEO variable suggests. Renewable energy, as being domestically produced and replenishable, is a critical vehicle to improve the energy independence of countries. Ultimately, we conclude that energy policy matters for the current account balance and subsequently for the well-being of OECD economies.

Since accumulation of renewable energy in a country takes time, an increase in the share of renewable-based sources in total electricity output is not significantly related to current account deficits in the short-term. The same argument is valid for the net energy import-current account deficit relationship. As the sources of energy utilized in a country are highly dependent on the energy policies of the country, and it definitely takes time to see the implications of these policies, net energy imports are not associated with the current account deficit in the short-term.

Finally, we also find strong evidence for the twin deficit hypothesis both in the short-run and long run. In other words, an increase in the level of budget deficit is associated with an upsurge in the current account deficit problem, causing deterioration. Furthermore, the causality test of [43] that considers the cross-sectional dependence among the countries in the sample reveals that there is bidirectional heterogeneous causality between current account deficit and budget deficit.

Group-based estimations also reveal interesting long-term results. Group 2 (the ones whose per capita incomes are over USD 25,000 but below USD 50,000) and Group 3 countries (the ones having more than USD 50,000 per capita income) do not differ in terms of the interaction of the current account deficit with either net energy import or renewable energy share in total electricity output, except the size of the coefficients. More specifically, the coefficients of energy-related indicators for Group 2 countries are higher than those of Group 3 ones, suggesting that energy policy matters more for Group 2 countries’ current account imbalances in the long-term. This might be attributed to the fact that Group 2 countries are more vulnerable to energy price shocks as their energy import dependencies are higher. Thus, even small improvements in the renewable energy landscape in these countries would help correct current account imbalances.

As for the interaction between current account deficit and fiscal deficit in the long-term, two country groups differ from each other. While in Group 2 countries we confirm the Ricardian view that postulates no relationship between current account deficit and fiscal deficit, Group 3 countries are in line with the Keynesian view that proposes a negative interrelation between the two.

Finally, short-run results of group-based panel estimations reveal that the only variable that is significantly associated with the current account deficit is the fiscal deficit. An increase in fiscal deficit worsens the current account deficit in both country groups.

7. Conclusions, Policy Recommendations and Limitations

7.1. Conclusions

In this study, we investigated the interrelation of current account deficit (CAB) as a percentage of GDP with selected indicators, namely, net energy import in total final energy consumption (NEI), the share of renewable energy in total electricity production (REN_TEO), and fiscal deficit as a percentage of GDP (FAB). Utilizing annual data in the period of 1990–2021, a Panel ARDL methodology is applied for 15 OECD countries.

The results of long-term estimations reveal that as net energy import increases, current account deficit deteriorates. The results of our study imply that countries that have a lower level of energy import dependency do not confront current account imbalances. On the other hand, in the case that countries use more of renewable energy in their total electricity output, their current account deficit levels improve in the long-term. However, in the short-term, the share of renewable-based sources in total electricity output is not significantly related to current account deficits. Thus, we acknowledge that it will definitely take time to accumulate adequate low-carbon energy capacity and see its positive implications on the economy. It is highly likely that global and country-based increases in renewable energy capacities in the last decade would support the economic well-being of many countries in the upcoming future.

In addition, we find strong evidence for the twin deficit hypothesis both in the short-run and long run. In other words, an increase in the level of budget deficit is associated with an upsurge in the current account deficit problem. This finding also implies that different economic policies might have close interactions with each other.

Finally, group-based estimations suggest that energy policy matters more for the OECD countries whose income levels are within the mid-range of the sample. Thus, we argue that even small improvements in the renewable energy landscape in these countries would help correct current account imbalances.

7.2. Policy Recommendations

There exists a large gap between global mitigation ambition and policy implementation. While mitigation ambition needs to be scaled up by all countries, differentiated responsibilities and conditions of advanced and developing countries should be accounted for. Domestic policy implementations should pace accordingly.

In this regard, to align with Paris Agreement goals, countries’ investment in clean energy is significantly outpacing spending on fossil fuels. Still, given the elevated energy prices and highly uncertain economic environment, global efforts to decarbonize the economy might not be too ambitious in the short-term. Yet, just for the sake of economic benefits, countries around the globe try to increase the share of renewables in their energy mix. As argued by [4], increased investment directed to more sustainable energy options is also resulting from the affordability and security concerns triggered by the global energy crisis. With the help of increased renewable capacities, countries that are highly dependent on energy imports might overcome challenges related to energy security and avoid negative economic consequences. Renewable energy as being domestically produced and replenishable would greatly support the energy trade balance of OECD countries. The results also imply that OECD countries should devise holistic economic policies that target simultaneously decreasing budget and current account deficits as the two move together.

If we open an additional parenthesis for Türkiye, we might definitely mention that Türkiye is not an exception. The interaction between investigated variables in Türkiye is found to be similar to other OECD countries in our sample.

As a country highly dependent on energy imports, the energy trade balance has been the main factor behind the current account deficit of Türkiye. Therefore, energy policies are crucial and have important implications for the Turkish economy as well. Increasing the share of renewables in total energy supply will gradually lead to a decrease in the net energy import of Türkiye and improve energy independence. Subsequently, increasing the share of renewable energy in total energy supply will lead to an increase in the share of renewables in total electricity output.

Fortunately, in line with its “More Domestic, More Renewable” energy strategy, Türkiye’s energy policy has given top priority to utilizing renewable energy sources to the maximum extent, while decreasing energy import dependency and improving supply security. The priority areas set in [44], the National Energy Plan (2020–2035), will not only help support decarbonization efforts but also be helpful in decreasing the current account deficit.

Notwithstanding the recent positive initiatives, such as diversifying the energy mix and installed capacity to the benefit of renewable sources, there is still room for improvement in Türkiye. Thus, a large amount of climate mitigation finance will need to be channeled to energy investments.

As a final note, Türkiye’s “More Domestic, More Renewable” energy policy aspirations will definitely support the development of domestic clean energy technologies, increase the value-added of certain products, and decrease energy import dependency. These improvements will eventually lead to a decrease in the energy trade balance and improve the current account balance of Türkiye.

7.3. Further Improvement Areas

Since energy trade balances are only available in annual frequency, this study utilizes annual data. Thus, there is room for further improvement of the study. For instance, future research might utilize different energy-related indicators that have monthly frequency. The research could also be replicated such that it includes different country groups in the analysis. On the condition that an adequate sample size is attained, additional variables might be included in the model. This would in turn enrich the discussions. Furthermore, the study might be replicated for individual countries rather than using a panel of countries. Finally, future research might propose additional insights by employing spatial econometric methods to capture the interdependencies among countries.