Evaluating Renewable Energy’s Role in Mitigating CO₂ Emissions: A Case Study of Solar Power in Finland Using the ARDL Approach

Abstract

This study examines Finland’s increasing investment in solar energy as part of its broader strategy to transition to renewable energy sources. Despite its northern location and limited sunlight during winter months, Finland has effectively harnessed solar power, especially during its long summer days. We conducted a PESTLE analysis, highlighting political commitment to climate agreements, economic benefits of solar energy, public support for clean energy, technological advancements, regulatory support, and environmental benefits. In addition, we applied the autoregressive distributed lag model (ARDL) to study the impact of GDP per capita, urbanization (URB), foreign direct investments (FDI), and the share of primary energy consumption from solar (SOL) on ${\mathrm{C}\mathrm{O}}_2$ emissions for Finland during 1990–2022. The long-term findings show that SOL, URB, and FDI negatively impact ${\mathrm{C}\mathrm{O}}_2$ emissions, while GDP positively impacts ${\mathrm{C}\mathrm{O}}_2$ emissions. Solar energy deployment reflects the effectiveness of solar energy as a cleaner alternative to fossil fuels. Urbanization also shows a negative impact on ${\mathrm{C}\mathrm{O}}_2$ emissions due to better infrastructure and more efficient energy use in urban areas. FDI suggests that investments are oriented towards sustainable technologies. Industrial activity associated with economic expansion may indicate the positive effect of GDP in ${\mathrm{C}\mathrm{O}}_2$ emissions. Although economic growth has typically resulted in increased emissions, strategic investments in renewable energy and sustainable urban development can greatly reduce this effect. Policy initiatives in Finland should concentrate on promoting green investments and enhancing urban planning with a focus on environmental sustainability.

1. Introduction

Renewable energy is essential to the shift to a low-carbon economy in the face of growing obstacles to both economic growth and environmental sustainability [1]. Recognized for its dedication to environmental regulations and renewable energy innovation, Finland provides an ideal setting to explore the impact and viability of solar energy in reducing ${\mathrm{C}\mathrm{O}}_2$ emissions.

Finland faces significant geographical challenges in harnessing solar energy [2] due to its northern latitude, leading to long, dark winters and short daylight hours during the winter months. Despite these challenges, the summer months offer extended daylight, which can be effectively used for solar power generation. Currently, the solar energy capacity in Finland is growing, although it still represents a small fraction of the total energy mix. The Finnish government has encouraged solar energy adoption by offering various incentives, including financial support like feed-in tariffs and grants, to facilitate its integration into the national grid. Finland primarily relies on other renewable energy sources, particularly hydro and wind power, with solar energy seen as a complementary source that can help balance the energy mix.

While Finland faces geographical challenges like Norway and Iceland, its policy support and growing capacity are helping to elevate the role of solar energy. Sweden and Denmark have more favorable conditions and stronger solar sectors, driven by robust policy incentives and better solar radiation. Solar energy in Finland is emerging as a complementary energy source, supported by government policies, despite the significant geographical challenges.

Undertaking research on the influence of solar energy on ${\mathrm{C}\mathrm{O}}_2$ emissions with a particular focus on Finland would provide a number of distinct problems [3,4]. In the context of wellbore stability and geological formation analysis, an appropriate strength criterion is essential, as highlighted by [5]. The Mohr–Coulomb criterion, which has been adjusted to account for the impacts of hydrate, emphasizes the significance of accurate analytical techniques for wellbore stability in particular geological conditions. This could have an impact on the application and assessment of energy technologies in Finland.

Additionally, the challenges associated with hydraulic fracturing technology, such as reservoir destruction and water pollution, highlighted by [6], emphasize the difficulties encountered when attempting to apply alternative methods like ${\mathrm{C}\mathrm{O}}_2$-based fracturing fluid. These aspects indicate the complexity of managing emissions and adopting new technologies in sensitive environments, which is also relevant for exploring renewable energy sources in specific climatic conditions, such as those in Finland. Finland’s northern location results in significant variations in daylight hours throughout the year. This means that during the long, dark winter months, solar energy generation is limited, impacting its overall contribution to the energy mix and its potential for reducing ${\mathrm{C}\mathrm{O}}_2$ emissions. Harsh winter conditions, including heavy snowfall and low temperatures, might have an impact on solar panels’ effectiveness. Snow cover can reduce solar panel output, posing operational challenges and affecting data consistency over time. Compared to countries with more favorable solar conditions, such as those closer to the equator or with sunnier climates, Finland has lower solar irradiance levels. This limits the potential energy generation from solar installations and influences the economic viability of large-scale solar projects. Finland’s adoption of solar energy is relatively recent compared to other European countries. The limited historical data make it challenging to conduct long-term studies on the impact of solar energy on ${\mathrm{C}\mathrm{O}}_2$ emissions.

Economic factors such as upfront costs, return on investment, and the availability of financing options influence the rate of solar energy deployment. Finland’s economic conditions and market dynamics may present unique challenges compared to other countries with different economic landscapes.

The research employs both qualitative and quantitative analyses to comprehensively assess the impact of solar energy on ${\mathrm{C}\mathrm{O}}_2$ emissions in Finland. Qualitative analyses, such as SWOT or the analysis of political, economic, social, technological, legal, and environmental factors (PESTLE) are recommended by the Environmental Impact Assessment (EIA) and International Organization for Standardization regulation (ISO 50001:2018, Geneva, Switzerland) frameworks for analyzing the impact of renewable energy [7]. Also, another study also uses PESTLE analysis to examine the sustainable development of solar energy potential in Turkey [8]. Our study leverages PESTLE analysis to identify specific opportunities and challenges within the Finnish context, such as government support, upfront costs, social acceptance, technological innovations, legal frameworks, and environmental impacts.

The primary goal of the study is to evaluate the contribution of solar energy in reducing ${\mathrm{C}\mathrm{O}}_2$ emissions in Finland, using the ARDL model to analyze the long- and short-term relationships between GDP per capita, renewable energy consumption, urbanization, foreign direct investment, and ${\mathrm{C}\mathrm{O}}_2$ emissions. Through this analysis, the study aims to provide practical recommendations for policy makers and the private sector, aiming at an efficient transition to more sustainable energy sources and a significant reduction in greenhouse gas emissions.

By combining PESTLE analysis with the ARDL model, the study provides a holistic perspective on the factors influencing the deployment and efficiency of solar energy in Finland. This integration allows for a novel understanding of how qualitative factors interact with economic and energy dynamics, ensuring that our recommendations are both contextually relevant and economically sound.

Our study is structured in several sections. Section 2 covers a review of the literature on the most important scientific studies thar are relevant to our study’s primary goal, and Section 3 presents the methodology of the PESTLE analysis, the ARDL model, and the data collection stage. The empirical results are presented in Section 4, and the conclusions, research limitations, and suggested recommendations are presented in Section 5.

2. The Stage of Knowledge in the Field

In recent decades, the world has witnessed a significant shift towards renewable energy sources in response to growing concerns about climate change [9,10] and the harm that carbon emissions cause to the environment [11]. This change was driven by the need to reduce dependence on fossil fuels, which are the main contributors to global greenhouse gas emissions [12]. In this context, solar energy has become one of the most promising solutions for a sustainable energy future [13]. Due to its widespread availability and renewable nature, solar energy offers significant opportunities to reduce ${\mathrm{C}\mathrm{O}}_2$ emissions [14,15,16], thereby contributing to global efforts to combat climate change.

The investigation of how solar energy affects ${\mathrm{C}\mathrm{O}}_2$ emissions is of major importance in the current context, where the international community is increasingly concerned with finding sustainable solutions to limit global warming. Solar energy, one of the renewable energy solutions, is seen as a key solution in reducing dependence on fossil fuels and reducing ${\mathrm{C}\mathrm{O}}_2$ emissions worldwide. By capturing solar energy and turning it into electricity, countries can significantly reduce greenhouse gas emissions, thereby helping to meet the goals set out in worldwide accords such as the Paris Agreement [17]. In addition, the widespread adoption of solar energy drives technological innovation [18,19], creates jobs in new sectors [20,21], and supports local economies [22], making it an essential component of sustainable energy strategies.

2.1. Literature Review of Qualitative Studies on Renewable Energy

Through the in-depth investigation of complex phenomena, qualitative research offers a thorough comprehension of how solar energy affects carbon emissions. Also, qualitative studies allow detailed exploration of the perspectives and motivations of actors involved in the shift towards renewable energy sources [23,24]. Through interviews, case studies, and other qualitative methods, researchers can identify perceived barriers and opportunities in the adoption of solar technologies, thereby providing valuable information that can guide implementation strategies and policies.

Guangul et al. [25] carried out a study in which they applied qualitative SWOT analysis to assess the use of suitable solar energy at the minimum cost and in the environment. Another study [26] focuses on renewable energy sources, with an emphasis on solar energy sources in Romania, considering that Romania has the technological potential and the geographical positioning suitable for the use and production of such energy. The authors of the study applied SWOT analysis to highlight the technological status and prospects for the development of renewable energy.

Another aspect analyzed through qualitative techniques is the sustainability of carbon capture and storage (CCS) technologies from a multidisciplinary perspective [27]. The case study in [27] was conducted for Finland, and the authors used the PESTLE framework to analyze the most significant drivers and barriers impacting these CCS technologies in Finland, as well as the general environment. Shasavary and Akabari [16] summarize the advantages of using solar energy and address the obstacles to its broad implementation. Salam and Khan [28] examine the factors driving Saudi Arabia’s shift towards solar energy as an alternative source of energy. The use of solar applications, particularly photovoltaic (PV) systems, is considered the most effective and affordable way to provide basic energy services in the region. By offering solar-powered home systems for indoor lighting, this can be encouraged. In the event that trustworthy and comparable quantitative datasets are unavailable, Holma et al.’s [29] qualitative assessment framework may be utilized to evaluate the environmental effects of producing renewable energy. It is a viable source of information about the effects of renewable energy. Another study [30] uses a qualitative method to compare Indonesia’s carbon tax policies with those of Sweden and Finland, as well as to analyze the reasoning behind the country’s carbon cost laws. According to the study’s authors, the carbon tax will finance research into renewable energy sources, finance the development of environmentally friendly technologies, and provide financial support for green enterprises as society moves toward a low-carbon future. In

Table 1. Comparison of previous qualitative research.

First Author, Year, Ref.	Objective	Methods	Main Findings
Guangul, 2019, [25]	Assessment of the advantages, disadvantages, possibilities, and threats related to solar energy utilization.	SWOT Analysis	Strengths: Solar energy is environmentally friendly, has low long-term costs, and is versatile. Opportunities: Technological advancements will reduce costs over time, making solar energy more accessible. Weaknesses: Significant obstacles include low efficiency, high starting costs, and the requirement for energy storage. Threats: Transitioning from fossil fuels to solar energy in the industry is difficult due to existing utilities that operate on conventional fuels.
Lupu, 2016, [26]	The authors assess Romania’s renewable energy sector’s potential, present situation, and future perspective, with a particular emphasis on solar energy.	SWOT Analysis	The regulatory framework pertaining to renewable energy sources is vital to the sector’s development, since it must be both viable and effective. Increasing public knowledge and support for renewable energy initiatives requires educating investors, developers, decision-makers, and the general public about the advantages and opportunities that come with using solar energy.
Pihkola, 2017, [27]	Sustainability analysis of carbon capture and storage (CCS) technologies from a multidisciplinary perspective. Also, another objective of this study is to identify the main factors and barriers influencing the implementation of CCS technologies in Finland and the impact on the general environment.	PESTLE Analysis	The PESTLE study revealed important legal and economic constraints that limit the advancement and application of CCS technologies. In Finland, carbon capture and utilization, or bio-CCS, may be a practical alternative to the production of energy from fossil fuels.
Shahsavari, 2018, [16]	Benefits of solar energy utilization and barriers to widespread adoption.	Literature review	Identifies technological and economic barriers to solar energy adoption
Salam, 2018, [28]	Factors driving Saudi Arabia’s shift towards solar energy.	Case study—qualitative analysis	Photovoltaic systems are an effective way to provide basic energy services.
Kumar, 2023, [31]	Evaluation of the actual status, importance, availability, and applications of solar technologies in Indian agriculture. The study also analyzes the socioeconomic and environmental impact, economic benefits, strengths, weaknesses, and future technological potential of solar energy in agriculture.	SWOT analysis; Economic analysis; Literature review; Environmental Impact Assessment.	Solar technologies can significantly reduce farmers’ dependence on conventional energy sources, thus contributing to the reduction in CO2 emissions. Solar energy can bring important economic benefits to farmers, reducing long-term energy costs. There is great potential for implementing solar technologies in various agricultural operations in India. It is anticipated that future technology developments will decrease solar energy’s cost and increase its efficiency, making it more feasible and cheaper for broad usage in agriculture.
Mohtaram, 2024, [32]	The study explores biochar-based photocatalytic methods for CO2 conversion and hydrogen (H2) production in solar fuel generation, highlighting the importance of improving the efficiency and stability of these photocatalysts.	Evaluation of current approaches in CO2 conversion and associated strategies, including the use of solar concentrators and thermal methods to improve the generation of solar fuels by artificial photosynthesis.	The study highlights the need to advance biochar-based photocatalysts, highlighting the importance of improving their efficiency and stability. Emphasis is also placed on innovative design concepts and the exploration of scalable manufacturing techniques, aiming to advance solar fuel generation technology for practical applications.
Ahmed, 2024, [33]	The study assesses current electricity consumption in households in the Gulf Cooperation Council (GCC) and proposes three scenarios for reducing energy consumption and CO2 emissions.	SWOT analysis. PESTLE analysis.	In the first scenario, lighting is suggested to be provided by solar photo-voltaic (PV) panels or a hybrid solar and wind PV system. This might result in significant energy savings, as lighting accounts for 8% to 30% of the electricity consumed in GCC houses. The second scenario considers the replacement of conventional appliances with energy-efficient appliances, which use about 20% less energy. The third scenario focuses on influencing consumer behavior towards sustainable energy consumption.
Thompson, 2023, [34]	In order to slow down climate change, the study investigates the role that electricity plays in the switch to renewable energy sources.	Systematic analysis of literature review. PESTLE analysis.	Smaller renewable technologies are developing, but transition is hindered by stranded assets, high upfront costs, variability of solar and wind power, infrastructure, difficulty in decarbonizing transport and industry, material resource constraints, and fossil fuel support. The analysis discovered that although the amount of renewable electricity has increased in 2021, renewable energy sources are not yet replacing fossil fuels; rather, they are creating new energy demands. As a result, greenhouse gas emissions have increased.

, several articles containing quality analyses were highlighted in which we identified the main objectives, the applied qualitative methods, and the main findings.

2.2. Literature Review of Quantitative Studies on Renewable Energy

Quantitative research is important for providing measurable evidence of the effect of carbon emission from solar energy. By using statistical and econometric models, these studies can quantify the impact of using solar energy on lowering greenhouse gas emissions, allowing for a precise and rigorous assessment of ecological and economic benefits.

Numerous studies take into account and examine how solar energy affects carbon emissions. Ajanovic [35] estimated that although renewable fuels produce lower ${\mathrm{C}\mathrm{O}}_2$ emissions compared to gasoline, they face economic limitations. Similarly, a study by Dimitrijevic and Dimitrijevic [36] discovered ${\mathrm{C}\mathrm{O}}_2$ emissions drop when the share of renewable energy sources in an energy mix increases.

Using the quantile-on-quantile method, Sharif et al. [37] investigated the dynamic link between solar energy consumption and ecological footprint for the top 10 nations with the highest solar energy consumption from 1990 to 2017. Their findings revealed that solar energy consumption leads to the lowering of the ecological footprint. Güney and İnce [38] use OLS, FMOLS, and CCEMG models to study the long-term link between solar energy and ${\mathrm{C}\mathrm{O}}_2$ emissions for a panel of 26 countries during 2000–2019. The authors recommend that governments provide subsidies to companies that use solar energy in the production process. This approach will lead to reduced carbon emissions, fostering a greener environment.

In China, Zhou et al. [39] investigated the uses of solar and wind energy in two scenarios: with and without legislation aimed at mitigating climate change. According to estimates, China’s power sector’s share of solar and wind energy will drop by over 15% and over 10%, respectively. The research by Güney [40] uses data from 35 countries with varying wealth levels between 2005 and 2018 to investigate the effects of solar energy and governance on ${\mathrm{C}\mathrm{O}}_2$ emissions. When cross-sectional dependence and slope uniformity are taken into consideration, the research shows that using more solar energy dramatically lowers ${\mathrm{C}\mathrm{O}}_2$ levels. Similarly, improved governance also negatively impacts ${\mathrm{C}\mathrm{O}}_2$ emissions. The combined effect of solar energy and governance further enhances this reduction.

In papers by Oprea and Bâra [41,42,43] the authors examined on-grid and off-grid photovoltaic (PV) systems of varying sizes and introduce a reliable PV forecasting method. The strategy used deterministic and stochastic models, machine learning, deep learning, and weather data-driven feature engineering to create a hybrid meta-learning forecasting technique. According to Chandrasekharam and Ranjith Pathegama [44], countries manufacturing solar PV cells will emit significant amounts of ${\mathrm{C}\mathrm{O}}_2$, other than the emissions from thermal power plants that use coal. Although solar PV cells do not emit ${\mathrm{C}\mathrm{O}}_2$ during electricity generation, their lifetime emissions are substantial.

Table 2. Comparison of previous quantitative research.

First Author, Year, Ref.	Objective	Methods	Main Findings
Al-Janabi, 2023, [45]	Evaluation and development of a software model to maximize solar energy production using advanced prediction techniques and multi-parameter objective functions.	Deep learning method (DMP-DGBM).	The DMP-DGBM model provides accurate results with an R2 of 0.9742, an MSE of 0.0099, and an RMSE of 0.0522. The implementation of the model takes only 80 milliseconds, demonstrating its efficiency and superior performance in predicting the solar peak.
Soto, 2024, [46]	Evaluating the carbon reduction impact of solar energy integration into specific U.S. electric grids.	Scenario analysis	The study found that the installation of solar energy facilities significantly reduced carbon dioxide emissions; the biggest reductions were observed in the Southwest and California as a result of the integration of solar plants. In order to reach emission reduction targets, it was determined that the New York and New England regions would need to implement additional policy measures and incentives.
Kinnunen, 2024, [47]	Evaluating the Environmental Phillips Curve Hypothesis in Finland’s STIRPAT 1990–2022 to understand the balance between economic growth and environmental impact.	Autoregressive distributed lag (ARDL)	The study demonstrates that, whereas GDP and urbanization raise GHG emissions, renewable energy use sharply lowers these emissions. Short-term data support the idea that urbanization has little effect on greenhouse gas emissions; this may be because Finland has highly developed urban planning. The study highlights the value of renewable energy in mitigating environmental effects and offer essential data to help policymakers strike a balance between environmental sustainability and economic growth.
Güney, 2024, [38]	The study examines the long-term relationships between solar power, globalization, coal power consumption, economic growth, and CO2 emissions in 26 countries over the period 2000–2019.	The method of correlated effects on mean groups (CCEMG); The OLS, FMOLS, and CCEMG estimates.	The use of solar energy lowers CO2 emissions; for every 1% increase in solar energy consumption, CO2 emissions are reduced by 0.0106671%. Long-term CO2 emissions and solar energy consumption are causally related in both directions. Globalization has no significant effect on CO2 emissions; instead, rising coal power use and economic expansion are the main drivers of CO2. The study shows how solar energy is a powerful tool for decreasing carbon emissions. Environmental sustainability requires promoting solar energy use and supporting investments in this area. Greener environments will result from reducing carbon emissions.
Perone, 2024, [48]	The study examines the long-term relationship between decoupled renewable energy production and carbon dioxide (CO2) emissions per capita for a group of 27 OECD countries over the period 1965–2020.	Panel-autoregressive distributed lag (ARDL) models.	It was discovered that the sources of CO2 emissions—hydroelectricity, nuclear, wind, solar, and aggregate geothermal and biomass (GEOB)—correlated adversely and considerably. It has been demonstrated that solar energy, hydroelectricity and geothermal energy are the most efficient renewable resources for lowering CO2 emissions. The Granger non-causality demonstrated unidirectional causality between CO2 emissions and nuclear power, bidirectional causality between CO2 and biofuel and GEOB, and unidirectional causality from solar, wind power, and hydroelectricity. The findings held true for a range of model assumptions and recommended an expedited shift to GEOB, hydroelectricity, and solar power in OECD nations in order to lower CO2 emissions and enhance environmental sustainability.

organizes the extensive data provided across various studies, presenting a comparison of their objectives, methods, and main findings.

3. Methodology and Data Collection

In this study, we propose to use two complementary methods—PESTLE analysis and the ARDL model—to evaluate the contribution of solar energy to Finland’s reduction of ${\mathrm{C}\mathrm{O}}_2$ emissions.

Figure 1 presents the methodological flow that we applied in this study. After the setting of the main research objective and a thorough investigation of the existing literature, the PESTLE analysis was performed and will be detailed in this section. The next step was to collect data relevant to the set objective. In order to understand how the selected indicators have evolved, summary statistics were performed. Once the data were understood, we started building the ARDL model following the steps in the flow presented.

Using Our World in Data as a source, we collected data on selected key economic indicators. In the next step, we computed descriptive statistics to understand the underlying characteristics and assessed skewness and kurtosis to check the normality of the data distributions. In the next step, we started building the ARDL model. We applied the Phillips–Perron test to determine the order of integration of each variable and used the unit root test with structural breaks to account for possible changes in the data and to identify the mix of variables I (0) and I (1). The ARDL model was selected by the lag VAR selection criterion and by using the Akaike information criterion. Next, the Bayer–Hanck cointegration test was applied to assess the presence of long-run relationships between the variables, and we performed the ARDL bounds test to confirm cointegration between the variables. Subsequently, long-run coefficients were estimated using the selected ARDL model and we analyzed the error correction model for short-run dynamics and speed of adjustment. After establishing the model and its coefficients, diagnostic and stability tests were performed. We also tested its robustness by using the alternative methods FMOLS, DOLS, and CCR. In addition, in order to explore the direction and temporal relationship between variables, we applied the Granger causality test.

The PESTLE framework [49] is a strategic tool used to analyze the macro-environment of a business, divided into six segments: political, economic, social, technological, legal, and environmental. This methodical technique helps to diagnose the external environment, identifying threats and opportunities that can enhance market competitiveness [50,51]. Interactions between governments, countries, or other political actors determine the political factor. Basically, this factor determines how government actions influence the economy or an economic system. Different political situations such as economic crises, political stability, or internal political situations characterize external political factors.

The economic factor includes key actors in the economic environment whose actions have both long- and short-term effects. Interest rate, inflation, and unemployment are economic factors that can be analyzed from this perspective. Also, certain costs can affect, from an economic point of view, subsidies as well. Regarding the social component, it includes a social environment analysis of the market being studied and assesses factors including training, population awareness cultural trends, demography, and social acceptance. The technological factor is directly dependent on technological trends and involves automation, research and development, as well as the degree of technological awareness that a market possesses. The legislative or legal factor includes laws, regulations, and norms that may affect the economic system in question. The environmental or ecological factor analyzes and measures the impact that environmental changes have on the analyzed systems, but also in inverse relation, i.e., the impact of the analyzed systems on the environment. This factor includes events such as natural disasters, recyclability, climate change, resource utilization, and resource availability.

The variables that have been chosen for the time-series analysis that follows, together with their sources, are listed in

Table 3. Variables specification.

Variables	Acronym	Measurement Unit	Source
${\mathrm{C}\mathrm{O}}_2$ emissions per capita	${\mathrm{C}\mathrm{O}}_2$	tonnes	Our World in Data
Foreign direct investments, net inflows	FDI	% of GDP	Our World in Data
Gross domestic product	GDP	constant 2015 $USD	Our World in Data
Share of primary energy consumption from solar	SOL	%	Our World in Data
Urbanization	URB	%	Our World in Data

for the years 1990–2022. Given the abbreviations used in our study, to provide clarity for the reader, all abbreviations utilized in this study have been listed in Appendix A,

Table A1. Abbreviations.

Acronym	Definition
SWOT	Strengths, Weaknesses, Opportunity, and Threats
PESTLE	Political, Economic, Social, Technological, Legal, Environmental
ARDL	Autoregressive Distributed Lag
FDI	Foreign Direct Investments
GDP	Gross Domestic Product
SOL	Share of primary energy consumption from solar
URB	Urbanization
ECM	Error Correction Model
ECT	Error Correction Term
VAR	Vector Autoregression
FPE	Final Prediction Error
AIC	Akaike Information Criterion
SC	Schwarz Information Criterion
HQ	Hannan–Quinn Information Criterion
PP	Phillips–Perron
FMOL	Fully Modified Ordinary Least Squares
DOLS	Dynamic Ordinary Least Squares
CCR	Canonical Cointegrating Regression

. The impact of GDP, SOL, URB, and FDI on $\mathrm{C}\mathrm{O}2$ emissions in Finland during 1990–2022 will be under analysis.

The relationship between the variables is as follows:

$$\mathrm{C}\mathrm{O}2=a_0+a_1{GDP}_t+a_2{SOL}_t+a_3{URB}_t+a_4{FDI}_t+{\epsilon }_t$$

(1)

The time series for $\mathrm{C}\mathrm{O}2$, GDP, and SOL have been converted to natural logarithms in order to mitigate sudden fluctuation in the data [52].

Equation (1) is transformed into an ARDL (n, p, q, r, s) regression:

$$\begin{gathered} ∆{\mathrm{C}\mathrm{O}2}_t=a_0+{\sum }_{k=1}^n{a}_{1,k}{∆CO2}_{t-k}+{\sum }_{k=1}^p{a}_{2,k}{∆GDP}_{t-k}+{\sum }_{k=1}^q{a}_{3,k}{∆SOL}_{t-k}+ \\ {\sum }_{k=1}^r{a}_{4,k}{∆URB}_{t-k}+{\sum }_{k=1}^s{a}_{5,k}{∆FDI}_{t-k}+{\lambda }_1{CO2}_{t-1}+{\lambda }_2{GDP}_{t-1}+{\lambda }_3{SOL}_{t-1}+ \\ {\lambda }_4{URB}_{t-1}+{\lambda }_5{FDI}_{t-1}+{\epsilon }_t \end{gathered}$$

(2)

$∆$ represents the first difference, n, p, q, r, and s are the lag orders, and the $a_{1,k},a_{2,k},\dots ,a_{5,k}$ denote that each differenced term has a unique coefficient depending on its lag $k$. The Bayer and Hanck [53] cointegration test provides robust results by integrating four distinct cointegration techniques: Engle and Granger [54], Johansen [55], Boswijk [56], and Banerjee et al. [57], abbreviated as EG, J, BO, and BA, respectively. It utilizes Fisher F-statistics to prove cointegration. The formulations of the test, following the Fisher method, are given by Equations (3) and (4):

$$EG-J=-[\ln \left(PEG\right)+\ln \left(PJ\right)]$$

(3)

$$EG-J-BO-BA=-2[\ln \left(PEG\right)+\ln \left(PJ\right)+\ln \left(PBO\right)+\ln (PA)$$

(4)

The test probabilities for the EG, J, BO, and BA tests are represented by the symbols PEG, PJ, PBO, and PA in Equations (3) and (4). The null hypothesis of no cointegration is rejected if the calculated Fisher statistic is greater than the complex value established by Bayer and Hanck [53]. The study’s results are corroborated using the ARDL bounds testing approach from Pesaran et al. [58] and the approach of Granger causality. Using F-statistics, this approach compares the option of cointegration to the null hypothesis of no cointegration. When the F-statistic is greater than the upper critical bound (I (1)), cointegration is confirmed and the null hypothesis is rejected. Equation (5) formulates the error correction model (ECM) in the presence of cointegration.

$$\begin{gathered} ∆{CO2}_t=a_0+{\sum }_{k=1}^n{a}_{1,k}{∆CO2}_{t-k}+{\sum }_{k=1}^p{a}_{2,k}{∆GDP}_{t-k}+{\sum }_{k=1}^q{a}_{3,k}{∆SOL}_{t-k}+ \\ {\sum }_{k=1}^r{a}_{4,k}{∆URB}_{t-k}+{\sum }_{k=1}^s{a}_{5,k}{∆FDI}_{t-k}+{\Gamma ECM}_{t-1}+{\epsilon }_t \end{gathered}$$

(5)

The notation $a_{1,k},a_{2,k},\dots ,a_{5,k}$ ensures that each differenced lag term has its own coefficient, reflecting the proper statistical treatment of each lag in the ARDL model. It is necessary for the error correction term (ECT) to be statistically significant, negative, and not more than −2 [59].

The normality test, the ARCH test, the Breusch–Godfrey serial correlation test, the LM test, and the Ramsey RESET test were performed. The cumulative sum (CUSUM) and cumulative sum of squares (CUSUMSQ) tests proved the model’s stability [60]. The CUSUM test is primarily used to detect gradual changes or drifts in the mean level of a process or time series over time. It is useful for identifying changes in the regression coefficients of a model. CUSUM calculates the cumulative sum of deviations from the mean of the model’s residuals (errors). This means that for each time point, the difference between the observed and expected values is added. CUSUMSQ is used to detect sudden or significant changes in the variance of a data series or process. It is more sensitive to detecting abrupt changes than the simple CUSUM. The CUSUMSQ chart compares the cumulative sums of the squares of the residuals with a reference line. Any significant deviation from the reference line indicates a possible structural change.

Advanced econometric techniques, such as Fully Modified Ordinary Least Squares (FMOLS), Dynamic Ordinary Least Squares (DOLS), and Canonical Cointegrating Regression (CCR), are utilized to estimate long-run connections among integrated variables that become stationary when they are differenced. These techniques are crucial for cointegration analysis, which examines long-term equilibrium relationships between non-stationary time series.

FMOLS, developed by Phillips and Hansen [61], corrects for serial correlation and endogeneity in cointegrated systems, making adjustments to the OLS estimator for more reliable long-run parameter estimates, especially useful in small samples. DOLS, proposed by Stock and Watson [62], addresses serial correlation and endogeneity by including leads and lags of the first differences in the independent variables, resulting in consistent and efficient long-term coefficient estimates, and is easy to implement with good finite-sample properties.

As part of our methodology, we included the Granger causality test to investigate the dynamic relationships between the economic variables analyzed. This analysis allows us to determine the direction and temporality of short-term [63] influences between variables such as solar energy consumption, GDP, and ${\mathrm{C}\mathrm{O}}_2$ emissions. The Granger causality test is essential to identify whether changes in one variable can predict changes in another variable, providing additional information to the long-run analysis performed by cointegration models. If one time series can aid in the prediction of another time series, it is tested for Granger causality. If there is a causal relationship between X and Y, then X contributes to making Y’s prediction better. On the other hand, Y influences the prediction of X if Y causes X. It is indicated by bidirectional causality that predictions are influenced by both X and Y. The general formulations of the Granger test for pairs, as proposed by Granger [64], are presented in relations (6) and (7) [65]:

$$Y_t={\sum }_{i=1}^n{\alpha }_i{Y}_{t-1}+{\sum }_{j=1}^n{\beta }_j{X}_{t-j}+{\mu }_{1t}$$

(6)

$$X_t={\sum }_{i=1}^n{\omega }_i{X}_{t-1}+{\sum }_{j=1}^n{\delta }_j{Y}_{t-j}+{\mu }_{1t}$$

(7)

4. Empirical Results

4.1. Qualitative Analysis: PESTLE Framework

PESTLE (political, economic, social, technological, legal, and environmental) analysis is a useful tool for understanding the external environment in which a particular technology or industry operates. In the context of our study on evaluating the role of solar energy in reducing ${\mathrm{C}\mathrm{O}}_2$ emissions in Finland, the PESTLE analysis gives us an overview of the factors that can influence the implementation and efficiency of solar energy.

Finland offers a perfect environment for investigating the effects and practicality of solar energy because of its reputation for environmental policies and renewable energy advancements. It is essential to investigate the political, economic, social, technological, legal, and environmental aspects that may have an impact on the process of lowering ${\mathrm{C}\mathrm{O}}_2$ emissions and switching to more sustainable energy sources. In this analysis from

Table 4. PESTLE analysis of the role of solar energy in reducing ${\mathrm{C}\mathrm{O}}_2$ emissions in Finland.

Political	Economic	Social	Technological	Legal	Environment
Government support: Finland has strong policies to support renewable energy, including subsidies and incentives for solar installations.	Upfront costs: Installing solar panels involves high upfront costs, but long-term costs are reduced due to low maintenance and savings on energy bills.	Public awareness: Increasing awareness among the population about the benefits of renewable energy and the impact on the environment.	Innovation and development: Technological progress in the efficiency of solar panels and the development of new materials to increase efficiency and durability.	Favorable legislation: Laws that support the installation of solar panels and provide tax breaks for the adoption of renewable energy.	Emission reduction: Solar energy contributes significantly to the reduction of CO2 emissions, supporting efforts to combat climate change.
International agreements: Joining the Paris Agreement and other international commitments requiring a clear target for reducing CO2 emissions.	Job creation: The development of solar energy creates employment opportunities in the solar system installation and maintenance sector.	Social acceptance: Community support and acceptance of solar technologies as viable energy solutions.	Integration with existing grids: Need for infrastructure development to effectively integrate solar energy into existing power grids.	Norms and standards: Clear regulations on the quality and safety of solar installations to ensure compliance and optimal performance.	Protection of natural resources: The use of solar energy reduces the pressure on natural resources and contributes to the preservation of the environment.
Regulations: Strict regulatory policies on CO2 emissions and promoting the use of green energy to meet climate target.	Energy independence: Increasing the use of solar energy can reduce dependence on fossil fuel imports, stabilizing the economy in the long term.	Education and training: The need for educational programs to train solar energy specialists and promote widespread adoption.	Storage solutions: Advances in energy storage technologies are essential to manage the intermittency of solar power and ensure a steady flow of power.	Contractual agreements: The need for strong contracts for the supply of solar energy and ensuring a legal framework for investment.	Impact on biodiversity: Solar installations must be carefully planned to minimize impact on natural habitats and biodiversity.

, we will identify and evaluate the key factors in each PESTLE category, highlighting both the advantages and challenges associated with the use of solar energy in Finland.

In terms of the political factor, the Paris Agreement [66], effective since 2016 and ratified by Finland, holds significant importance. This agreement is an important political element as it aims for long-term climate change adaptation. The climate commitments outlined in the Paris Agreement have been in effect since 2020. Other regulatory policies aimed at managing the reduction of ${\mathrm{C}\mathrm{O}}_2$ emissions and promoting green energy in Finland include climate neutrality by 2035 [67], renewable energy integration [68], support for innovation [69,70], and the increased adoption of electric vehicles [71]. Besides political and regulatory commitments, political and industry co-operation with neighboring countries can affect public perceptions and help find best practices.

From the perspective of the economic factor, installing solar panels in Finland involves significant upfront costs, which include purchasing the panels, installation, and additional equipment like inverters and mounting systems. Market demand and public interest draw economically from competitive cost-effective technological solutions. A household solar panel system can cost anywhere from EUR 9979 to EUR 12,197 on average [72], depending on the installation’s complexity and size. However, Finland offers subsidies and tax incentives to reduce these initial costs. For example, households can apply for investment grants that cover 40% of the installation costs, making solar energy more accessible [73].

In Finland, the transition to renewable energy, especially solar electricity, has a favorable impact on job creation. The renewable energy sector, including solar energy, is expected to create thousands of jobs over the next decade [74]. These jobs span various fields such as manufacturing, installation, maintenance, research and development, and energy consultancy [20]. Promoting the use of solar energy contributes to Finland’s energy independence. Additionally, by raising the proportion of locally generated renewable energy, Finland reduces its reliance on imported fossil fuels. This shift enhances national security by mitigating the risks associated with energy supply disruptions and price volatility in the global market. Further, energy independence supports sustainable economic growth by keeping energy expenditures within the country and fostering technological innovation [75]. These aspects coincide also with the social factor in the PESTLE analysis, e.g., on the public awareness and equitable access to solar energy.

Technological factors are directly related to trends in technology and the possibility of its use in the analyzed environment. According to this perspective, Finland is renowned for its technological advancements and for being at the forefront of numerous technological advancements on a worldwide scale. Finland’s education system prioritizes technology- and digital-based instruction, guaranteeing a high degree of digital literacy across the nation. This includes advances in mobile technology, software advancements, security, and data analytics. However, Finland faces a shortage of skilled technology workers, especially in the areas of digitalization, artificial intelligence (AI), and data analytics [76]. Finland has made significant investments in its energy infrastructure, which have made it easier to integrate renewable energy sources into the country’s current electrical systems. This includes solar technologies that successfully integrate into national power grids, enabling an easier transition to more sustainable energy sources. With a developed infrastructure and policies to support energy innovation, Finland can support the large-scale adoption and deployment of solar technologies [76,77].

Regarding the legal factor, there are several laws and regulations to promote green energy and reduce ${\mathrm{C}\mathrm{O}}_2$ emissions in Finland. For example, the legal act no. 1396 on renewable energy sources of 2010 provides for various subsidies and financial incentives to produce renewable energy, including solar energy [78]. The Electricity Market Act 588/2013 [79] specifies the conditions for the efficient national and regional electricity market and was drafted with the aim of providing consumers with a guarantee of security of electricity supply as well as competitive prices. This has been addressed by the Law on amending the Electricity Market Act 497/2023 [80], which applies to retailers regarding electricity suppliers. In addition, the European standards adopted by Finland SFS-EN 61215 [81] and SFS-EN 61730 [82] set performance and safety requirements for photovoltaic modules. These include tests for durability, efficiency, and resistance to varying environmental conditions. The standards developed by the Finnish Solar Energy Association [83] specify the technical requirements for the installation and maintenance of solar systems, thus ensuring optimal performance and maximum safety, while mitigating the impact on a legally robust framework. In addition, Long-Term Power Purchase Agreements (PPAs) [84] with comprehensive clauses on feed-in rates, contract durations, and dispute resolution procedures are critical for the provision of solar energy. They provide legal and financial certainty for investors and producers.

In terms of the environmental factor, the Greenhouse Gas Emission Reduction Strategy 2050 has been developed, which includes specific measures to reduce ${\mathrm{C}\mathrm{O}}_2$ emissions by promoting solar and other renewable energy sources. Targets include reducing emissions by 80–95% by 2050 compared to 1990 levels [85].

According to Finland’s Ministry of Environment [86], any large-scale solar energy project must go through an Environmental Impact Assessment (EIA) process to assess the impact on biodiversity and natural habitats. This process ensures that projects are carefully planned and implemented to minimize negative environmental impacts. These concrete laws and rules underline Finland’s commitment to promoting solar energy and reducing ${\mathrm{C}\mathrm{O}}_2$ emissions, while protecting the environment and biodiversity.

The comprehensive PESTLE analysis conducted provides valuable insights into the multifaceted factors impacting the execution and effectiveness of solar energy in reducing ${\mathrm{C}\mathrm{O}}_2$ emissions in Finland. By examining political, economic, social, technological, legal, and environmental aspects, we identified both opportunities and challenges associated with the adoption of solar energy. Key political drivers, such as Finland’s commitment to international climate agreements and supportive legislation, create a favorable environment for solar energy development. Economically, while high upfront costs pose challenges, government subsidies and job creation in the renewable energy sector offer significant advantages. Social acceptance and technological innovation further enhance the potential for solar energy integration. However, legal frameworks and environmental considerations necessitate careful planning to ensure sustainable development.

4.2. Quantitative Analysis: ARDL Approach

The ARDL model analysis complements the PESTLE findings by quantitatively assessing the impact of key economic indicators such as GDP, solar energy consumption, urbanization, and foreign direct investment on ${\mathrm{C}\mathrm{O}}_2$ emissions in Finland.

The yearly evolution of Finland’s five chosen key indicators from 1990 to 2022 is depicted in Figure 2. The ${\mathrm{C}\mathrm{O}}_2$ emissions show a decreasing trend over time. There is notable variability in the early years, with peaks and troughs, but overall, there is a significant downward trend. The GDP shows a clear increasing trend over the period. There are some fluctuations, but the overall trajectory is upward, indicating economic growth. The usage of solar energy starts at a very low level and remains almost flat until around year 25, after which it shows a sharp increase. This indicates a recent and rapid adoption of solar energy in recent years. The urbanization rate shows a steady increase over the entire period. The increase is gradual but consistent, indicating continuous urban development. The FDI shows high variability with no clear long-term trend. There are significant ups and downs throughout the period, indicating fluctuations in foreign investment.

In the case of Finland, the trend in ${\mathrm{C}\mathrm{O}}_2$ emissions reflect several economic, political, and technological factors that have influenced emission levels during the analyzed period. As shown in Figure 2, the data indicate a general decline in ${\mathrm{C}\mathrm{O}}_2$ emissions from 1990 to 2022, with some fluctuations during certain periods. This is due to the fact that Finland has implemented strict environmental policies and promoted the use of renewable energy sources, which have contributed to the reduction in ${\mathrm{C}\mathrm{O}}_2$ emissions [87,88]. This includes the adoption of solar, wind, and other clean energy sources [89,90]. Additionally, increased urbanization and technological innovations have led to more efficient resource management and the development of technological solutions that reduce emissions.

Descriptive statistics for the variables after logarithmic transformation are shown in

Table 5. Summary statistics.

	${\mathbf{C}\mathbf{O}}_2$	GDP	SOL	URB	FDI
Mean	2.35	10.56	$-$5.70	4.41	1.08
Median	2.43	10.63	$-$6.35	4.41	1.55
Maximum	2.63	10.75	$-$1.14	4.45	2.37
Minimum	1.87	10.23	$-$7.78	4.37	$-$1.31
Std. Dev.	0.20	0.17	1.83	0.02	1.05
Skewness	$-$0.92	$-$0.66	1.28	$-$0.03	$-$0.86
Kurtosis	2.73	1.99	3.67	1.94	2.65
Jarque–Bera	3.90	3.14	7.95	1.25	3.53
Probability	0.14	0.20	0.01	0.53	0.17

. ${\mathrm{C}\mathrm{O}}_2$ emissions and GDP are moderately left-skewed, indicating that most values are higher but with some low outliers. SOL is highly right skewed with heavy tails, indicating a few very high values. URB and FDI are nearly symmetric and exhibit distributions close to normal. URB and FDI data are less variable, while GDP and ${\mathrm{C}\mathrm{O}}_2$ emissions show moderate variability. SOL has high variability and significant deviations from normality.

All variables are integrated in order 1, according to the results of the Phillips–Perron (PP) unit root test [91] (see

Table 6. PP unit root test results.

Variable	Level	First Difference	Order of Integration
	T-Statistics	T-Statistics
${\mathrm{C}\mathrm{O}}_2$	0.76 (0.95)	$-$73.45 *** (0.00)	I (1)
GDP	$-$0.99 (0.93)	$-$4.22 ** (0.01)	I (1)
SOL	$-$0.40 (0.98)	$-$7.89 *** (0.00)	I (1)
URB	$-$2.42 (0.35)	$-$3.69 ** (0.03)	I (1)
FDI	$-$2.83 (0.20)	$-$9.32 *** (0.00)	I (1)

**, *** indicate the significance of variables at 5% and 1% levels, respectively.

).

Structural break unit root tests are particularly valuable in time series analysis due to their ability to account for sudden shifts in data that traditional unit root tests like PP might miss. Therefore, we further apply the Vogelsang and Perron [92] break point unit root test.

Table 7. Vogelsang and Perron breakpoint unit root test results.

Variables	Level		First Difference		Order of Integration
	T-Statistics	Break Year	T-Statistics	Break Year
${\mathrm{C}\mathrm{O}}_2$	−2.58 (0.87)	2011	−8.15 *** (0.00)	2003	I (1)
GDP	−3.79 (0.23)	1996	−7.11 *** (0.00)	2008	I (1)
SOL	−2.90 (0.73)	2014	−6.59 *** (0.00)	2015	I (1)
URB	−8.73 *** (0.00)	2009	−4.31 * (0.07)	2011	I (0)
FDI	−7.11 *** (0.00)	1997	−11.81 *** (0.00)	1998	I (0)

*, *** indicate the significance of variables at 10% and 1% levels, respectively.

indicates that ${\mathrm{C}\mathrm{O}}_2$, GDP, and SOL are I (1), while URB and FDI are I (0). Since we have a combination of I (0) and I (1) orders of integration, the ARDL model will be used.

For the vector autoregression (VAR) model, choosing a lag order of 1 is the best option, according to three of the five criteria in

Table 8. VAR lag order selection criteria.

Lag	LogL	LR	FPE	AIC	SC	HQ
0	77.14	N/A	$1.42\times {10}^{-10}$	$-$8.48	$-$8.24	$-$8.46
1	228.90	196.40 *	$5.54\times {10}^{-17}$ *	$-23.40$	$-$21.53 *	$-$23.25
2	261.98	23.34	$6.27\times {10}^{-17}$	$-$24.35 *	$-$21.65	$-$24.08 *

* indicates the lag order selected by the criterion; LR: sequential modified LR test statistic (each test at 5% level); FPE: Final prediction error; AIC: Akaike information criterion; SC: Schwarz information criterion; HQ: Hannan–Quinn information criterion.

.

For lag selection in the optimal ARDL model, we also used the Akaike information criterion, evaluating the top ten models based on AIC values, as shown in Figure 3. The ARDL (1,1,1,1,0) has the lowest AIC value, indicating that it provides the best fit for our data, taking into account the complexity penalty.

The obtained model is ARDL (1,1,1,1,0). Boswijk [56] and Banerjee et al. [57] propose tests based on error correction models, which may be more appropriate in such contexts. Even in asymptotic terms, there is no test that is universally stronger, thus choosing the right one becomes difficult (see Elliott et al. [93]). When one test rejects the null hypothesis while another does not, complexity develops, which makes it harder to interpret the results. In addition, the p-values of these tests are not perfectly correlated (Gregory et al. [94]), implying that we cannot rely solely on the test with the lowest p-value. Ignoring the multiple-testing aspect would result in an over-biased test. The tests’ power is additionally impacted by the fact that imperfect correlation between them suggests that they are not equal. Pesavento [95] demonstrates that the power classification of cointegration tests depends heavily on the value of a perturbation parameter, particularly the long-run correlations of the error terms driving the variables. This has led to the development of the Bayer–Hanck [53] cointegration test, which integrates several tests to obtain more robust results in the context of mixed variables. Therefore, in our study, we opted for the Bayer–Hanck test to assess cointegration, given its ability to better handle situations with mixed variables I (0) and I (1). This test provides a more integrated and robust approach to evaluate cointegration relations under the particular conditions of our data.

Table 9. Bayer–Hanck cointegration test.

Tests	Engel–Granger EG	Johansen J	Banerjee BA	Boswijk BO
Test statistic	$-$4.02	43.35	$-$4.16	26.97
p-value	0.11	0.00	0.03	0.00
EG-J	16.02	5% critical value, 10.57
EG-J-BA-BO	35.55	5% critical value, 20.14

indicates that the F-statistic values computed by the EG-J and EG-J-BA-BO methods exceed the critical values at the 5% significance level. This outcome supports the rejection of the null hypothesis of no cointegration at the 5% level.

Table 10. Results of ARDL cointegration bounds test.

Tests	Value	K (Number of Regressors)
F-statistic	4.92	4
Critical value bounds
Significance	I (0)	I (1)
10%	2.20	3.09
5%	2.56	3.49
1%	3.29	4.37

displays the cointegration bounds test results for the ARDL model.

Table 10 indicates that the calculated F-statistic is 4.92, which exceeds the upper critical bound I (1). These findings imply that the variables included in the model have an equilibrium connection over the long term. This indicates that the variables are cointegrated, meaning that over the long term, they tend to move together despite short-term volatility. This result is important for economic policy analysis as it suggests that changes in variables such as GDP, solar energy consumption, and foreign direct investment have lasting effects on ${\mathrm{C}\mathrm{O}}_2$ emissions in Finland. The corresponding long-term coefficients are presented in

Table 11. Long-run estimated results.

Variables	Coefficient	T-Statistics	Prob.
GDP	0.99	3.88	0.00 ***
SOL	$-$0.09	$-$4.79	0.00 ***
URB	$-$5.01	$-$1.76	0.09 *
FDI	$-$0.07	$-$2.18	0.04 **
C	13.58	1.25	0.22

*, **, *** indicate the significance of variables at 10%, 5%, and 1% levels, respectively.

.

Table 11 indicates that in Finland, a 1% rise in GDP results in a 0.99% long-term increase in ${\mathrm{C}\mathrm{O}}_2$. In Finland, there is a clear and substantial correlation between economic growth and ${\mathrm{C}\mathrm{O}}_2$ emissions, as evidenced by this nearly proportionate ratio. Higher ${\mathrm{C}\mathrm{O}}_2$ emissions result from increased economic activity as Finland’s economy expands. Increased energy use, transportation, industrial output, and other economic activities that generally lead to higher greenhouse gas emissions could be the cause of this. Policymakers must strike a balance between environmental sustainability and economic prosperity. To disentangle economic growth from ${\mathrm{C}\mathrm{O}}_2$ emissions, some strategies to consider are energy efficiency improvements, investments in cleaner technology, and the promotion of renewable energy sources.

In Finland, a 1% increase in SOL results in a 0.09% long-term drop in ${\mathrm{C}\mathrm{O}}_2$. By dispensing with more carbon-intensive energy sources like coal, oil, or natural gas, increasing the usage of solar energy contributes to a reduction in ${\mathrm{C}\mathrm{O}}_2$ emissions. The overall carbon footprint of energy production is decreased by this move toward cleaner energy. The promotion of solar energy as a realistic method for lowering greenhouse gas emissions is supported by this research. By offering incentives, subsidies, regulatory support, and financial investments in solar infrastructure, policymakers can promote the use of solar energy.

In Finland, a 1% increase in URB results in a long-term 5.01% drop in ${\mathrm{C}\mathrm{O}}_2$. Significant drops in ${\mathrm{C}\mathrm{O}}_2$ emissions are correlated with higher levels of urbanization. Public transit is frequently better developed in urban areas, which can lessen the need for private vehicles and cut emissions associated with transportation. Stricter environmental laws and efficiency requirements may be implemented by cities, which would reduce emissions from industry and buildings. In metropolitan regions, higher population densities have the potential to reduce per capita emissions and promote more efficient energy use. Improving urban infrastructure and promoting urbanization can be useful tactics for lowering ${\mathrm{C}\mathrm{O}}_2$ emissions. To optimize these advantages, policymakers should concentrate on encouraging public transit, improving energy efficiency in urban environments, and designing sustainable urban layouts.

${\mathrm{C}\mathrm{O}}_2$ decreases by 0.07% in the long run with every 1% rise in FDI. Foreign investors often bring advanced technologies and practices that are more energy-efficient and environmentally friendly. FDI can stimulate the adoption of green technologies and renewable energy projects. Foreign companies may be subject to stricter environmental standards from their home countries, which they implement in their operations abroad, leading to lower emissions.

Table 12. ECM model for short-run estimated results.

Variables	Coefficient	T-Statistics	Prob.
D(GDP)	2.61	4.86	0.00 ***
D(SOL)	$-$0.20	$-$4.21	0.00 ***
D(URB)	$-$14.85	$-$2.69	0.01 ***
CointEq (−1)	$-$0.83	$-$6.18	0.00 ***
R-squared	0.66
Adjusted R-squared	0.61

*** indicates the significance of variables at 1% level.

presents the ECM model for short-term effects, which is essential for understanding the impact of the variables on ${\mathrm{C}\mathrm{O}}_2$ emissions. An ECT value of −0.83 indicates that each period corrects about 83% of the long-term equilibrium divergence. This implies a relatively fast adjustment speed, with most of the adjustment happening within the first period after a shock. A high absolute value (close to 1) of the ECT, like −0.83, indicates that the system adjusts quickly to changes, meaning that any short-term shocks to the system (such as sudden changes in GDP, solar energy usage, urbanization, or FDI) will dissipate quickly, and the dependent variable will return to its long-term equilibrium state. This fast adjustment process can be beneficial as it shows resilience in the system, suggesting that Finland’s policies and economic structures are effective in maintaining equilibrium in response to changes in the explanatory variables.

In the short term, ${\mathrm{C}\mathrm{O}}_2$ emissions rise by 2.61 units for every unit increase in GDP, as seen in Table 12. Increased energy use, industrial activity, and other economic activity may be responsible for this, as these factors usually result in larger emissions. In the short term, ${\mathrm{C}\mathrm{O}}_2$ emissions are reduced by 0.20 units for every unit rise in SOL. Solar energy is a clean and renewable source of energy, and its increased usage helps to displace fossil fuel consumption, thus reducing emissions. In the short term, ${\mathrm{C}\mathrm{O}}_2$ emissions are significantly reduced by 14.85 units for every unit rise in URB rate. Given that urbanization is frequently linked to higher energy usage and emissions, this may seem in opposition. However, it could reflect a situation where urbanization in Finland is accompanied by improved infrastructure, better energy efficiency, and more stringent environmental regulations, leading to lower emissions. It might also reflect a shift towards more sustainable urban planning and public transportation systems that reduce the carbon footprint. The absence of a short-term effect of FDI on ${\mathrm{C}\mathrm{O}}_2$ emissions can be explained by several factors. FDI typically involves long-term investments in infrastructure, technology, and business operations. The effects of these investments on production processes, energy consumption, and emissions may not be immediate. It often takes time for the capital investments to be fully operational and to start influencing environmental outcomes. If FDI is concentrated in sectors that do not have an immediate impact on ${\mathrm{C}\mathrm{O}}_2$ emissions, such as services or technology sectors, the short-term effect on emissions might be negligible. Sectors like renewable energy or technology might even offset potential increases in emissions due to their cleaner nature. Finland has stringent environmental regulations and policies aimed at controlling emissions. These regulations apply to both domestic and foreign enterprises. Therefore, incoming FDI might already be aligned with these regulations, preventing any immediate spike in emissions. FDI in Finland may involve the transfer of green technologies and practices, which are designed to be environmentally friendly. The initial implementation phase of such technologies might not show immediate changes in emissions levels but could lead to long-term reductions.

4.2.1. Diagnostic and Model Stability Assessment

In this section, we focus on the diagnostic and stability analysis of the econometric model used in the study. The purpose of these tests is to check the validity of the model and to ensure that the model is correctly specified and robust. Correct diagnostics allow us to identify and correct possible problems such as serial correlation, heteroscedasticity, abnormal distribution of residuals, or possible model misspecification.

To assess these issues, we centralized a couple of statistical tests that we applied in our study, including the serial correlation test, the ARCH test for heteroscedasticity, the Jarque–Bera test for normality of the distribution of residuals, and the Ramsey RESET test for detecting model specification errors. In addition, we used CUSUM and CUSUM of Squares plots to analyze the stability of model parameters over time. These analyses are essential to ensure that the conclusions drawn from the model are reliable and relevant for the economic policies analyzed.

Table 13. Outcomes of stability and diagnostic testing.

Diagnostic Test	${\mathit{H}}_0$	Decision Statistics [p-Value]
Serial Correlation	There is no serial correlation in the residuals	Accept $H_0$ 0.87 [0.43]
Heteroscedasticity (ARCH)	There is no autoregressive conditional heteroscedasticity	Accept $H_0$ 0.01 [0.90]
Jarque–Bera	Normal distribution	Accept $H_0$ 1.15 [0.56]
Ramsey RESET	Absence of model misspecification	Accept $H_0$ 0.38 [0.70]

displays the diagnostic and stability test null hypotheses.

Figure 4 and Figure 5 display a red dashed line indicating that the CUSUM and CUSUM of Squares routes are still within the 5% significance criterion. They verify that the parameter of the model is stable.

4.2.2. Robustness Checks and Granger Causality Analysis

In this section, we address two key issues for validating and deepening our econometric analysis: robustness checks and Granger causality analysis. Robustness checks are essential to ensure that our model results are stable and do not depend on particular model specifications. To this end, we use alternative methods such as FMOLS (Fully Modified Ordinary Least Squares), DOLS (Dynamic Ordinary Least Squares), and CCR (Canonical Cointegrating Regression), which allow us to test the consistency of the estimates under various scenarios.

Granger causality analysis also provides additional information about the dynamic relationships between the variables under study, regarding the direction and temporality of the influences. The Granger causality test allows us to determine whether one variable can be considered as a short-run causal factor for another variable in the economic system analyzed. By combining these analyses, we ensure that our conclusions are both robust and comprehensive, providing a sound basis for policy recommendations.

FMOLS, DOLS, and CCR methodologies validate the precision of the statistical analysis of the research variables, especially considering the previously established cointegration relationships. The FMOLS, DOLS, and CCR long-term coefficients’ signs match those from the ARDL model in Table 11, as

Table 14. FMOLS, DOLS, and CCR long-term coefficients.

Variables	FMOLS Coefficient, (t-Statistics), [p-Value]	DOLS Coefficient, (t-Statistics), [p-Value]	CCR Coefficient, (t-Statistics), [p-Value]
GDP	1.02	0.88	1.00
	(4.72)	(3.34)	(4.55)
	[0.00] ***	[0.00] ***	[0.00] ***
SOL	−0.09	−0.10	−0.09
	(−3.81)	(−4.69)	(−4.28)
	[0.00] ***	[0.00] ***	[0.00] ***
URB	−6.44	−4.89	−6.69
	(−2.26)	(−1.92)	(−2.94)
	[0.03] **	[0.06] *	[0.00] ***
FDI	−0.03	−0.01	−0.03
	(−1.63)	(−0.58)	(−1.12)
	[0.12]	[0.56]	[0.22]
C	2.03	14.06	20.14
	(1.74)	(1.43)	(2.27)
	[0.10]	[0.16]	[0.03] **

*, **, *** indicate the significance of variables at 10%, 5%, and 1% levels, respectively.

illustrates. The only difference is the lack of long-term significance of FDI on ${\mathrm{C}\mathrm{O}}_2$, a result also confirmed by Georgescu and Kinnunen [96]. This difference highlights the importance of using multiple methods to verify results, ensuring a comprehensive understanding of the relationships between variables. Finland has robust environmental regulations and policies aimed at controlling ${\mathrm{C}\mathrm{O}}_2$ emissions. These regulations apply to all companies, including those receiving foreign direct investments. Foreign companies must comply with these stringent standards, which might mitigate any potential increase in emissions. The type of FDI Finland attracts might be concentrated in sectors that are less carbon-intensive, such as technology, services, and high-tech manufacturing. These sectors typically have lower ${\mathrm{C}\mathrm{O}}_2$ emissions compared to heavy industries. Finland’s economy is highly developed and characterized by a strong emphasis on sustainability and environmental protection. The structural aspects of its economy might inherently support lower emissions, regardless of FDI levels.

To test whether there is a significant causality between the analyzed variables, we conducted the Granger causality test (see

Table 15. Granger causality.

Null Hypothesis	F-Statistic	Prob.	Conclusion
GDP d.n.G.c. CO2	2.19	0.13
CO2 d.n.G.c. GDP	1.77	0.18
SOL d.n.G.c. CO2	3.18	0.05 *	$SOL\to {CO}_2$
CO2 d.n.G.c. SOL	1.60	0.22
URB d.n.G.c. CO2	3.03	0.06 *	$URB\to {CO}_2$
CO2 d.n.G.c. URB	0.52	0.59
FDI d.n.G.c. CO2	0.16	0.85
CO2 d.n.G.c. FDI	0.56	0.57
SOL d.n.G.c. GDP	0.24	0.78
GDP d.n.G.c. SOL	0.25	0.77
URB d.n.G.c. GDP	1.91	0.16
GDP d.n.G.c. URB	0.46	0.63
FDI d.n.G.c. GDP	0.67	0.51
GDP d.n.G.c. FDI	1.44	0.25
URB d.n.G.c. SOL	3.31	0.05 *	$URB\to SOL$
SOL d.n.G.c. URB	0.27	0.76
FDI d.n.G.c. SOL	0.42	0.65
SOL d.n.G.c. FDI	0.15	0.85
FDI d.n.G.c. URB	0.73	0.48
URB d.n.G.c. FDI	1.23	0.30

* indicates the significance of variables at 10% level; d.n.G.c. means “does not Granger cause”.

).

GDP does not Granger-cause ${\mathrm{C}\mathrm{O}}_2$. From an economic perspective, this indicates that ${\mathrm{C}\mathrm{O}}_2$ emissions have not been significantly influenced by growth or contraction in Finland’s economy over the studied period. This could imply that Finland’s economic activities have been relatively decoupled from ${\mathrm{C}\mathrm{O}}_2$ emissions, possibly due to effective environmental regulations, a shift towards a less carbon-intensive economy, or improvements in energy efficiency. Similarly, ${\mathrm{C}\mathrm{O}}_2$ emissions do not predict changes in GDP. This suggests that variations in ${\mathrm{C}\mathrm{O}}_2$ emissions, perhaps driven by energy consumption patterns or environmental policies, do not significantly influence economic growth. This might indicate a robust economy that is not heavily dependent on carbon-intensive industries.

The finding that solar energy uptake has a predictive relationship with ${\mathrm{C}\mathrm{O}}_2$ emissions implies that increases in solar energy usage are associated with subsequent reductions in ${\mathrm{C}\mathrm{O}}_2$ emissions. Economically, this highlights the effectiveness of renewable energy adoption in mitigating carbon emissions. It suggests that policies promoting solar energy contribute to reducing the carbon footprint in Finland, supporting the country’s goals towards sustainability and carbon neutrality. Changes in ${\mathrm{C}\mathrm{O}}_2$ emissions do not predict changes in solar energy adoption. This could mean that the adoption of solar energy technologies is more influenced by policy measures, technological advancements, and economic incentives rather than the direct pressure from increasing ${\mathrm{C}\mathrm{O}}_2$ emissions.

Urbanization Granger causing ${\mathrm{C}\mathrm{O}}_2$ emissions suggests that as urban areas expand or become more densely populated, ${\mathrm{C}\mathrm{O}}_2$ emissions tend to increase. Economically, this relationship might be due to increased energy demand, transportation needs, and consumption patterns associated with urban living. In order to reduce the negative effects of urban growth on the environment, the outcome highlights the necessity of sustainable urban design and green infrastructure. The lack of predictive power from ${\mathrm{C}\mathrm{O}}_2$ emissions to urbanization suggests that rising ${\mathrm{C}\mathrm{O}}_2$ levels are not a significant deterrent to urban expansion. This could indicate that other factors, such as economic opportunities, housing, and infrastructure development, primarily drive urban growth.

The lack of Granger causality in both directions (FDI and ${\mathrm{C}\mathrm{O}}_2$) implies no significant predictive relationship between FDI and ${\mathrm{C}\mathrm{O}}_2$ emissions. Economically, this might indicate that FDI in Finland has not been significantly associated with carbon-intensive sectors or that FDI inflows and outflows are balanced in terms of their environmental impact. This could reflect a diversified investment landscape where environmental concerns do not primarily drive FDI decisions, or vice versa.

The Granger causality test indicates that changes in SOL Granger cause changes in ${\mathrm{C}\mathrm{O}}_2$ emissions. This finding is consistent with the results obtained from the ARDL model in Table 11, which show that increases in solar energy consumption are associated with reductions in ${\mathrm{C}\mathrm{O}}_2$ emissions. This causal relationship supports the idea that expanding the share of solar energy in Finland’s energy mix directly contributes to lowering carbon emissions. The Granger causality result reinforces the understanding that solar energy plays a crucial role in reducing carbon emissions. As solar energy replaces fossil fuels in the energy mix, it helps decrease the overall carbon footprint, which is a key objective for Finland in meeting its climate goals. This relationship underscores the importance of continued investment in solar energy. Even if solar energy does not currently have a significant impact on GDP (as indicated by the lack of Granger causality between SOL and GDP), its role in reducing ${\mathrm{C}\mathrm{O}}_2$ emissions is clear and valuable. Policymakers should take this into account, emphasizing the environmental benefits of solar energy in their energy strategies. The causal link between SOL and ${\mathrm{C}\mathrm{O}}_2$ suggests that solar energy contributes positively to environmental sustainability. Over time, as the share of solar energy increases, its impact on reducing ${\mathrm{C}\mathrm{O}}_2$ emissions could become more significant, leading to broader environmental benefits. This justifies continued and even accelerated investment in solar technologies, which could also lead to future economic benefits as the technology matures and its adoption increases.

SOL does not Granger-cause GDP. This finding should be interpreted in the context of the current scale and efficiency of solar energy within Finland’s energy system. SOL has represented only a small fraction of Finland’s total primary energy consumption until recently; therefore, it might not yet have a significant impact on overall economic growth. In economies where solar energy forms a larger portion of the energy mix, the relationship between SOL and GDP might be more pronounced. Therefore, the current lack of Granger causality could reflect the fact that solar energy’s share is not yet large enough to drive noticeable changes in GDP. Even if SOL currently does not Granger-cause GDP, it does not mean solar energy lacks economic value. The strategic importance of transitioning to renewable energy sources, including solar, is critical for long-term sustainability. Investment in solar energy should continue, with a focus on increasing its share in the energy mix, improving technological efficiency, and ensuring better integration with other energy sources. Beyond immediate economic impacts, solar energy contributes to energy security and reduces environmental degradation. These benefits, although not directly reflected in GDP growth in the short term, are crucial for the long-term health of the economy and society.

5. Conclusions

Finland is renowned for its robust renewable energy laws, which include financial aid and incentives for solar panel installation. The Paris Agreement and other international commitments oblige the country to set clear targets for reducing ${\mathrm{C}\mathrm{O}}_2$ emissions. These political commitments create a favorable framework for the adoption and expansion of solar energy.

Several studies describe that although economic growth has typically resulted in increased emissions, strategic investments in renewable energy and sustainable urban development can greatly reduce this effect. For example, Bhowmik [97], Wang and Wang [98], and Dai et al. [99] discuss global strategies to decouple economic growth from carbon emissions, providing successful examples from Finland and other Nordic countries that have managed to combine economic growth with emission reductions by transitioning to renewable energy sources.

The PESTLE analysis has shown that, despite the high initial costs associated with solar panel installation, these are eventually covered by lower energy bills and fewer maintenance expenses. Finland offers subsidies and tax incentives to reduce these upfront costs, facilitating access to solar energy for households and businesses. Reducing reliance on imports of fossil fuels and promoting energy independence are two further benefits of the switch to renewable energy.

Public awareness of the benefits of renewable energy and its impact on the environment is also growing. Social acceptance and community support for solar technologies are key to the success of the energy transition. Education and training programs help to prepare solar energy specialists and promote widespread adoption.

The PESTLE analysis’s findings offer a thorough framework for comprehending the variables affecting Finland’s usage of solar energy and how those variables affect ${CO}_2$ emissions. Using the ARDL model, we were able to analyze the long-term relationships between ${\mathrm{C}\mathrm{O}}_2$ emissions per capita, foreign direct investment, GDP, solar energy consumption, and urbanization. This approach allowed us to identify the key variables that contribute to emission reductions and to assess the effectiveness of implemented policies and technologies. Thus, the PESTLE analysis and the ARDL model complement each other, providing an integrated perspective on the energy transition in Finland and its contribution to environmental sustainability.

According to the ARDL model, long-term increases in GDP of 1% result in increases in ${\mathrm{C}\mathrm{O}}_2$ emissions of 0.99%; increases in solar energy use of 1% result in reductions in ${\mathrm{C}\mathrm{O}}_2$ emissions of 0.09%; increases in urbanization of 1% result in reductions in ${\mathrm{C}\mathrm{O}}_2$ emissions of 5.01%; and increases in foreign direct investment of 1% result in reductions in ${\mathrm{C}\mathrm{O}}_2$ emissions of 0.07%.

In the short term, an increase in GDP is associated with a significant increase in ${\mathrm{C}\mathrm{O}}_2$ emissions, while an increase in the use of solar energy and the rate of urbanization leads to a significant decrease in ${\mathrm{C}\mathrm{O}}_2$ emissions.

These findings imply that a large-scale switch to solar energy can significantly lower Finland’s ${\mathrm{C}\mathrm{O}}_2$ emissions. But, in order to accomplish these objectives, a comprehensive strategy involving supportive legislation, financial support, societal acceptance, and technological innovation is required. The ARDL model’s conclusions underscore the necessity of striking a balance between environmental sustainability and economic growth, emphasizing the critical role that solar energy plays in this process. Although this aspect has been validated in some studies, we lack sufficient statistical evidence to assert concretely that there is a direct causal relationship between GDP growth during the analyzed period and solar energy. The Granger causality test results did not support this claim. Nonetheless, the findings demonstrated a strong causal relationship between the use of solar energy and the decline in ${\mathrm{C}\mathrm{O}}_2$ emissions, indicating that solar energy plays a vital role in environmental sustainability. Additionally, we discovered proof that urbanization promotes the use of solar energy and lowers ${\mathrm{C}\mathrm{O}}_2$ emissions. As with any scientific study, it is important to recognize potential limitations. The study focuses on data from 1990 to 2022. This time frame, while long, may not capture all long-term variations and may exclude recent events and policy changes that could influence ${\mathrm{C}\mathrm{O}}_2$ emissions and solar uptake. This limitation is also caused by the lack of complete data for the older period or for the last 2 years. The study results are also specific to Finland and may not be directly applicable to other countries or regions with different economic, political, and climatic characteristics. The impact of policies and infrastructure may vary significantly between countries, which limits the global applicability of the study findings. However, in terms of methodological flow and the use of analytical methods and the ARDL model, it can be replicated on different datasets specific to other countries. In addition, exogenous factors, such as economic crises, political changes, or natural disasters, can significantly affect ${\mathrm{C}\mathrm{O}}_2$ emissions and solar energy uptake, but cannot be predicted or fully modeled in the study. These limitations emphasize the need for a cautious approach in interpreting the results and formulating policies based on them.

In conclusion, policymakers should recognize that the system is responsive and adjusts quickly to changes in the independent variables. This rapid adjustment provides an opportunity for dynamic policy interventions since the system can quickly return to equilibrium after policy-induced shocks. Ensuring that policies do not cause excessive short-term disruptions is important, as the system’s quick adjustment means that any deviation (whether positive or negative) will be rapidly corrected.