Evaluating the Ecological Footprint of Biomass Energy: Parametric and Time-Varying Nonparametric Analyses

Abstract

The growing discourse surrounding biomass energy’s environmental ramifications has ignited debate among policymakers. While biomass remains a primary and readily accessible energy source, various studies have extensively examined its implications for health and the economy. However, there is a lack of evidence regarding its role in mitigating climate change. This study delves into the ecological footprint implications of biomass energy consumption in the Organisation for Economic Co-operation and Development (OECD) countries, spanning from 1990 to 2017. While the existing literature predominantly relies on parametric methodologies, offering estimates of biomass energy’s average impact on ecological footprints, it fails to capture temporal variations in this relationship. Consequently, this study employs both parametric and nonparametric time-varying techniques to elucidate the evolving impact of biomass energy utilization on ecological footprints across the studied nations. Findings from both analytical approaches converge to suggest that biomass energy usage amplifies the ecological footprint of OECD nations. Notably, the nonparametric analysis underscores the dynamic nature of this relationship over time. Based on these insights, policy recommendations are given to mitigate the adverse environmental consequences of biomass energy usage while exploring cleaner alternative energy sources.

1. Introduction

To achieve sustainability objectives, nations must effectively coordinate economic, environmental, and social facets. Environmental preservation has emerged as a paramount concern worldwide, given its profound impact on people across all regions, especially the marginalized and underprivileged. Environmental challenges such as severe weather events, air pollution, and global climate shifts disproportionately affect vulnerable populations [1], necessitating prompt action and strategic interventions. A promising avenue involves transitioning away from fossil fuels, which currently dominate 80% of global primary energy use and contribute 75% toward greenhouse gas emissions [2,3]. This transition involves embracing alternative energy sources like bioenergy, wind, hydro, geothermal, and solar power. As affirmed by the scientific community [2,3,4,5,6], renewable energy offers a pathway to curbing carbon emissions and fostering environmental well-being. Thanks to advancements in energy efficiency, technological innovation, and supportive policy frameworks, renewable energy adoption has experienced rapid growth in recent years [4].

Biomass energy stands as the leading renewable energy source in current usage. It accounted for around 10% of the total final energy consumption in 2020, which includes both traditional biomass (used mainly for cooking and heating in developing countries) and modern bioenergy (such as biofuels and biogas for electricity, heat, and transportation) [5,6]. Notably, modern bioenergy alone accounted for about 5% of global final energy consumption in 2020, representing a significant portion of renewable energy generation [4,6]. Projections indicate that biomass energy will maintain its significance in meeting global energy demands [7,8]. The International Energy Agency and World Bioenergy Association reported that bioenergy is the largest renewable energy source. As of the latest reports, bioenergy constitutes about 50% of the renewable energy mix and supplies around 10% of the world’s total final energy consumption [5,6]. The growing popularity of biomass energy can be attributed to its numerous advantages. Firstly, biomass energy offers versatility, serving various purposes such as cooking, electricity generation, heating, and transportation. Notably, it stands out as the sole renewable energy source directly convertible into liquid fuels. Secondly, biomass presents a cost-effective and straightforward production process, bolstering its appeal as a renewable energy option. By leveraging biomass energy, nations can reduce their dependency on fossil fuels while ensuring energy security [9]. Furthermore, biomass energy generation fosters employment opportunities, particularly in rural areas, thereby augmenting incomes and aiding in alleviating poverty [10]. Bioenergy has the potential to reduce greenhouse gas emissions compared to fossil fuels, particularly when it is sustainably sourced. Biomass combustion releases carbon dioxide, which is part of the current carbon cycle, offset by the carbon absorbed during the growth of the biomass. This is in contrast to fossil fuels, which release carbon that has been stored underground for millions of years, adding to the atmospheric carbon burden [5]. Moreover, bioenergy generally produces lower emissions of sulfur dioxide (SO₂), nitrogen oxides (NO_x), and particulate matter (PM) compared to coal and oil. These pollutants contribute to air quality problems such as acid rain and respiratory issues [11]. Therefore, bioenergy emits fewer pollutants and exerts a lower environmental impact compared to fossil fuels [5,12]; hence, biomass energy holds promise in curbing carbon emissions and combatting climate change on a global scale [7].

Moreover, biomass is recognized not only as a renewable energy source but also as a vital component in capturing and storing CO₂, thereby playing a crucial role in the carbon cycle. Beyond its use in energy production, biomass can be directly utilized in construction materials, paper production, and other applications where it effectively sequesters carbon for extended periods [13]. This dual function of biomass, providing both energy and long-term carbon storage, highlights its importance in mitigating climate change [14]. By integrating biomass into various sectors, we can enhance its contribution to reducing atmospheric CO₂ levels while simultaneously supporting sustainable development and environmental conservation efforts [15].

The future of biomass energy is undergoing a significant transformation due to the stringent climate goals established by the Paris Agreement, implemented in 2016, which replaced previous, more lenient regulations [16]. The Paris Agreement sets ambitious targets for reducing greenhouse gas emissions and transitioning to low-carbon energy sources, necessitating a reassessment of biomass energy’s role within this new framework [17]. As global policy shifts towards more rigorous climate action, future evaluations of biomass energy must address not only its renewable attributes, but also its full lifecycle emissions and environmental impacts [18]. Advances in technology and evolving regulatory standards will shape the sustainability of biomass energy, requiring continuous updates to its ecological footprint assessments to ensure alignment with international climate commitments [19]. Consequently, a comprehensive analysis integrating these dynamic factors will be crucial for optimizing biomass energy’s contribution to achieving global climate goals.

As biomass energy usage continues to proliferate, it has garnered significant attention from researchers. A substantial body of literature has focused on exploring its environmental ramifications [20,21,22,23], as well as its implications for economic growth [12,24,25,26], human health [27,28,29], and overall human development [30,31,32]. While much of this research has relied on carbon dioxide (CO₂) emissions as a primary indicator of environmental impacts, it has become increasingly clear that such a narrow focus may overlook the broader spectrum of environmental consequences [33]. Indeed, CO₂ emissions alone provide only a limited perspective on the overall environmental footprint of human activities.

Recognizing this limitation, researchers have turned to more comprehensive metrics, such as the ecological footprint (EF), pioneered by Wackernagel and Rees [34], to better assess the holistic impact of biomass energy utilization [35,36]. Unlike CO₂ emissions, the EF encompasses a wider array of factors, including cropland, grazing land, fishing grounds, built-up areas, forest cover, and carbon emissions. Moreover, the EF captures both the direct and indirect effects of human activities on the environment, providing a more nuanced understanding of environmental efficiency [37]. Thus, in assessing the environmental implications of biomass energy use, the EF emerges as a superior metric compared to the singular focus on CO₂ emissions.

Although the EF has emerged as a prominent metric in environmental research [38,39,40,41,42,43,44,45], empirical studies probing the nexus between biomass energy and ecological footprints remain relatively scarce [22,46,47]. This relationship has sparked debate among researchers, with divergent perspectives on the environmental impact of biomass energy use. While some argue for its environmental friendliness, citing reductions in greenhouse gas emissions [48], carbon emissions [7,49,50,51], and overall ecological footprints [52], others contend that biomass energy exacerbates carbon emissions and poses risks to the atmosphere and broader environment [22,23,53,54].

The existing literature predominantly employs parametric methodologies to measure the average impact of biomass energy use on ecological footprints, but fails to capture temporal dynamics in this relationship. To address this gap, our study utilizes both parametric and nonparametric time-varying techniques to assess the impact of biomass energy use on the ecological footprint of OECD nations over the time period of 1990 to 2017.

Our contribution to the academic discourse lies in the deployment of a generalized two-stage least square (G2SLS) random effects model to address endogeneity issues in estimating the impact of biomass energy use on ecological footprints. Additionally, we introduce a nonparametric panel data technique, employing the local linear dummy variable estimation method (LLDVE), to measure the time-varying effects of biomass energy use on ecological footprints. This novel approach overcomes the limitations of parametric estimates, which often obscure time-varying relationships between variables. The LLDVE technique captures ecological footprints by an unknown time trend, allowing for variations in the common trend function over the study period, and detecting common disturbances, such as shifts in government policies.

2. Literature Review

The significance of utilizing renewable energy to mitigate emissions and pollution has surged in response to the environmental challenges induced by climate change. Extensive research has investigated the environmental ramifications of diverse energy sources on carbon emissions, consistently demonstrating that renewable energy outperforms non-renewable sources in reducing carbon footprints [55,56,57,58,59,60,61,62,63]. Recent research by Dong et al. [64] and Cheng et al. [65] has specifically highlighted how renewable energy adoption can curb carbon emissions in BRICS countries. Similarly, various investigations have yielded comparable findings across different regions and nations, including studies on Sub-Saharan Africa’s top electricity producers [60], European Union member states [62], leading renewable energy adopters globally [56], the United States [61], Pakistan [66], and Italy [67].

Furthermore, scholarly attention towards understanding the environmental implications of biomass energy has markedly expanded in the past decade. A synthesis of the literature, elucidating the link between biomass energy utilization and environmental effects, is shown in

Table 1. A compilation of influential research discovering the connection between biomass energy usage and its environmental impact.

Authors	Years	Countries	Methods	Results
Danish [21]	1990–2014	26 OECD countries	PMG, FMOLS, DOLS	bio ↑ carbon emission
Mahmood et al. [23]	1980–2015	Pakistan	ARDL	bio ↑ carbon emission
Shahbaz et al. [68]	1980–2014	G7 countries	GMM	bio ↑ carbon emission
Solarin et al. [54]	1980–2010	80 countries	GMM, DCCE	bio ↑ carbon emission
Sinha et al. [69]	1990–2014	N-11 countries	GMM	bio ↑ carbon emission
Adewuyi and Awodumi [53]	1980–2010	West African countries	3SLS	bio ↑ carbon emission
Shah et al. [70]	1990–2017	38 Asian nations	CCEMG	bio ↑ carbon emission
Gao and Zhang [25]	1980–2010	13 Asian countries	FMOLS	bio ↑ carbon emission
Wang et al. [22]	1980–2016	G7 countries	DSUR	bio ↑ ecological footprint
Danish and Ulucak [71]	1982–2017	China	DARDL	bio ↓ carbon emission
Destek and Aslan [72]	1991–2014	G7 countries	AMG	bio ↓ carbon emission
Danish and Wang [20]	1992–2013	BRICS countries	GMM	bio ↓ carbon emission
Shahbaz et al. [51]	1990–2015	MENA countries	GMM	bio ↓ carbon emission
Shahbaz et al. [73]	1960–2016	USA	ARDL	bio ↓ carbon emission
Bilgili et al. [7]	1982–2011	USA	Hatemi-J casality	bio ↓ carbon emission
Dogan and Inglesi-Lotz [50]	1985–2012	22 countries	FMOLS	bio ↓ carbon emission
Katircioglu [74]	1980–2010	Turkey	ARDL	bio ↓ carbon emission
Bilgili [75]	1990–2011	USA	Hatemi-J causality	bio ↓ carbon emission
Baležentis et al. [48]	1995–2015	27 European Union nations	FE, FMOLS, PDOLS	bio ↓ GHG emission
Bilgili et al. [76]	1984–2015	USA	Wavelet coherence approach	bio ↑↓ carbon emission
Ahmed et al. [77]	1980–2010	24 European nations	PMG	bio /→ carbon emission
Sarkodie [47]	1970–2017	Australia	ARDL	bio /→ ecological footprint bio ↓ GHG emission
Kim et al. [78]	1973–2016	USA	DAG, ARDL	bio ↓ carbon emission

Note: ↑ = increase carbon emission, ↓ = reduce carbon emission, /→ = insignificant, bio = biomass energy consumption.

.

The academic discourse surrounding biomass energy’s impact on the environment remains unresolved, with prior studies failing to establish consensus. One faction argues that biomass utilization leads to diminished carbon emissions. For instance, Katircioglu [75] in Turkey concluded that biomass energy can contribute to CO₂ emission reduction. Similarly, Shahbaz et al. [73], using the ARDL method for the United States spanning from 1960 to 2016, found a decline in CO₂ emissions related to biomass energy. Comparable results were obtained by Bilgili et al. [76] in the United States, employing the wavelet coherence technique, and by Dogan and Inglesi-Lotz [50], who assessed the impact of biomass energy and income on carbon emissions in nations with significant biomass usage, advocating for biomass as a clean fuel. Additionally, Baležentis et al. [79] and Sarkodie et al. [47] highlighted a reduction in greenhouse gas emissions through biomass energy utilization, using panel data estimation techniques. Danish and Wang [20] recently investigated biomass’ environmental implications in BRICS countries, concluding that it diminishes pollution. Kim et al. [78] obtained similar results for the United States using an ARDL estimation method. Shahbaz et al. [51] emphasize that biomass energy consumption plays a crucial role in mitigating CO₂ emissions associated with foreign direct investment in Middle Eastern and North African countries. Likewise, Danish and Ulucak [71] reveal through dynamic autoregressive distributed lag simulations that biomass energy consumption in China significantly reduces CO₂ emissions. Similarly, Baležentis et al. [48] demonstrate that bioenergy contributes significantly to reducing greenhouse gas emissions in EU countries, employing an Environmental Kuznets Curve modeling approach. Moreover, Destek and Aslan [72] find that a 1% increase in disaggregated renewable energy consumption results in a notable 0.42% reduction in environmental pollution across G7 countries. Furthermore, Shah et al. [70] illustrate that biomass energy consumption significantly supports economic development in Asia, contingent upon the adoption of advanced and efficient technologies to optimize environmental benefits.

Conversely, another group of studies contends that biomass energy contributes to environmental pollution and heightened carbon emissions. Solarin et al. [54], employing the GMM technique across 80 nations, revealed a positive connection between biomass energy usage and CO₂ emissions. Correspondingly, Shahbaz et al. [68] found similar outcomes for G7 nations, while Adewuyi and Awodumi [53] reported increased carbon emissions from biomass energy in West African Nations, and Gao and Zhang [25] observed similar trends in 13 developing Asian countries.

The inconclusive nature of this debate underscores the necessity for further study of the association between biomass energy consumption and its ecological impact. Consequently, this research aims to diagnose the repercussions of biomass energy usage on ecological footprints by utilizing both parametric (G2SLS) and nonparametric time-varying (LLDVE) techniques. By leveraging these methods, the insights gained from this research endeavor will provide valuable guidance for policymakers seeking a deeper understanding of the biomass energy–environment nexus.

3. Data and Empirical Modeling

3.1. Data

We analyzed data spanning from 1990 to 2017 to inspect how biomass energy use affects the ecological footprints of OECD nations. Data on ecological footprints (measured in global hectares per capita) were sourced from the Global Footprint Network [80], while information on biomass energy consumption (measured in tons per capita) was obtained from the United Nations Global Material Flows Database, 2019. The KOF Globalization Index [81] provided data on globalization levels. Additionally, control variables were extracted from the World Bank database [82]. Digitalization (DT) was evaluated using a principal component analysis (PCA) score based on internet usage rates and mobile phone users (per 100 people).

Table 2. Detailed information on data.

Variable	Symbol	Definition	Measure	Source
Biomass energy consumption	BIO	The total amount of biomass energy consumption per population of a country	Tons per capita	Global Material Flows Database
Ecological footprint	EF	The ecological assets required to create the natural resources consumed	Gha per capita	Global Footprint Network
Digitalization (PCA score)	DT	Percentage of the total population is utilizing the internet	% of population	World Development Indicator
		The number of mobile cellular subscriptions per 100 individuals	Per 100 people	World Development Indicator
Natural resources	NR	The total natural resource rents comprise the combined value of oil, natural gas, coal (both hard and soft), mineral rents, and forest rents	% of GDP	World Development Indicator
Globalization Index	GI	Calculated based on economic flows and restrictions, information flows, personal contact, and cultural proximity	Index (2010 = 100)	KOF Globalization Index
Population density	PD	Measured by dividing the midyear population count by the land area in square kilometers	People/km2 of land area	World Development Indicator
Employment Rate	ER	The percentage of individuals aged over 15 who are employed	% of population	World Development Indicator

outlines the selected variables, their measurement units, and their respective sources.

Further, comprehensive summary statistics of the selected variables are presented in

Table 3. Summarized information of data.

Indicators	lnEF	lnBIO	lnDT	lnNR	lnGI	lnPD
Mean	1.664	1.163	0.000	−1.117	4.313	4.402
Median	1.693	1.181	0.723	−1.045	4.363	4.668
Minimum	0.552	0.173	−5.034	−7.037	3.729	0.798
Maximum	2.878	2.366	1.169	3.064	4.511	6.267
Std. Dev.	0.384	0.395	1.389	1.958	0.160	1.251
Skewness	−0.141	0.263	−1.325	−0.145	−1.323	−0.974
Kurtosis	4.034	3.492	3.688	2.350	4.316	3.773
Obs.	896	896	896	896	896	896

.

3.2. Econometric Model

This study utilizes a model, drawing upon insights from the literature review and previous studies [37,83], to assess the impact of biomass energy use on ecological footprints. Control variables such as natural resources, globalization, and digitalization are included in the analysis.

$$E{F}_{it}=f\left(BI{O}_{it},D{T}_{it},N{R}_{it},G{I}_{it},P{D}_{it},\right)$$

(1)

Equation (1) defines the variables as follows: EF for ecological footprint, BIO for biomass energy consumption, DT for digitalization, NR for natural resources, GI for the globalization index, and PD signifies population density. A simple multivariate model is employed to assess the effect of biomass energy usage on ecological footprints. Simultaneously, the natural logarithms of the series were considered to lessen variability and flatten the data. Equation (2) presents the log-linear version of the model.

$$\begin{gathered} Ln{\left(EF\right)}_{it}={\delta }_0+{\delta }_{1}ln{\left(BIO\right)}_{it}+{\delta }_{2}ln{\left(DT\right)}_{it}+{\delta }_{3}ln{\left(NR\right)}_{it}+{\delta }_{4}ln{\left(GI\right)}_{it}+ \\ {\delta }_{5}ln{\left(PD\right)}_{it}+{\epsilon }_{it} \end{gathered}$$

(2)

Here, “i” represents the studied nations, and “t” represents the years under consideration. The symbol ${\delta }_0$ signifies the intercept of the model, while ${\delta }_i$ (i = 1, 2,…,5) characterizes the coefficients of predictors, including BIO, DT, NR, GI, and PD. The term ε signifies the random error term affecting the ecological footprint. Given that this study incorporates panel data, an initial assessment is necessary to determine whether cross-sectional dependency exists.

3.2.1. Tests for Cross-Sectional Dependence and Slope Homogeneity

Globalization and economic integration can lead to spillover effects, where the determinants affecting one nation may influence neighboring countries as well. In panel data analysis, cross-sectional dependency can arise due to interrelationships between countries. One common flaw in earlier analytical methods is the assumption of cross-sectional independence. Research results obtained using such methodologies may be biased if cross-sectional dependence is overlooked [12,84]. We integrated cross-sectional dependence tests into our research to tackle this concern, employing the Breusch–Pagan Lagrange multiplier (LM) and Pesaran CD tests [85,86].

The LM test statistics are computed using the following equation:

$$L{M}^{BP}=T{\displaystyle \sum _{i=1}^{N-1}{\displaystyle \sum _{j=i+1}^N{\widehat{r}}_{ij}^2}}$$

(3)

When T becomes sufficiently large, the LM test may not be suitable. To address this limitation, the following CD test can be utilized as an alternative:

$$C{D}^P=\sqrt{{\displaystyle \frac{2T}{N(N-1)}}}\left({\displaystyle \sum _{i=1}^{N-1}{\displaystyle \sum _{j=i+1}^N{\widehat{r}}_{ij}^2}}\right)$$

(4)

In Equations (3) and (4), i and j denote the cross-sections (nations), T characterizes the year, N signifies the total number of cross-sections, and ${\widehat{r}}_{ij}^2$ denotes the correlation of the error terms.

Additionally, following the approval of cross-sectional dependence, we employed the Pesaran and Yamagata [87] slope homogeneity tests to evaluate the consistency of the slope coefficients. Previous analytical methods often overlooked country-specific characteristics, assuming homogeneity [88]. Pesaran and Yamagata [87] introduced a standardized dispersion test statistic ($\tilde{\mathrm{\Delta }}$) based on Swamy’s model [89] to scrutinize slope homogeneity.

$${\tilde{\mathrm{\Delta }}}_{adjusted}=N^{{\displaystyle \frac{1}{2}}}\left({\displaystyle \frac{\frac{1}{N}\tilde{S}-k}{\sqrt{2k}}}\right)$$

(5)

Here, $\tilde{S}$ represents the modified Swamy test. For small samples, Equation (5) can be rewritten as follows:

$${\tilde{\mathrm{\Delta }}}_{adjusted}=N^{{\displaystyle \frac{1}{2}}}\left({\displaystyle \frac{\frac{1}{N}\tilde{S}-k}{\sqrt{2k(T-k-1)/T+1}}}\right)$$

(6)

3.2.2. Generalized Two-Stage Least Square

To accommodate unobserved differences among countries, reflecting cross-sectional dependency, an additional random effect term ($u_i$) has been included in Equation (2). This yields the following random effect model:

$$\begin{gathered} \mathrm{l}\mathrm{n}{\left(EF\right)}_{it}={\delta }_0+{\delta }_1\mathrm{l}\mathrm{n}{\left(BIO\right)}_{it}+{\delta }_2\mathrm{l}\mathrm{n}{\left(DT\right)}_{it}+{\delta }_3\mathrm{l}\mathrm{n}{\left(NR\right)}_{it}+{\delta }_4\mathrm{l}\mathrm{n}{\left(GI\right)}_{it} \\ +{\delta }_5\mathrm{l}\mathrm{n}{\left(PD\right)}_{it}+u_i+{\epsilon }_{it} \end{gathered}$$

(7)

The key objective of this study is to assess the environmental repercussions of biomass energy utilization. Recognizing that biomass energy consumption could be endogenous, as it may be influenced by other unobserved variables, there’s a risk of downward bias in estimating the link between biomass energy use and ecological footprints. Hence, resolving the endogeneity issue is crucial for obtaining accurate, reliable, and efficient estimates. To tackle this, a G2SLS approach within a random effect model was employed. In this method, the employment rate (the proportion of employed individuals aged 15 and above) serves as an instrumental variable, potentially impacting biomass energy consumption. Although there’s no direct link between the employment rate and ecological footprints, an indirect relationship may exist via biomass energy usage. The two stages outlined in Equations (8) and (9) are employed to address potential endogeneity concerns in the study.

Reduced form equation (stage 1)

$${\ln \left(BIO\right)}_{it}={\beta }_0+\gamma {\ln \left(ER\right)}_{it}+\beta X_{it}+a_i+v_{it}$$

(8)

Structural equation (stage 2)

$${\ln \left(EF\right)}_{it}=\alpha +\delta {\ln \widehat{\left(BIO\right)}}_{it}+\eta X_{it}+u_i+{ϵ}_{it}$$

(9)

In the model, $\alpha$ and ${\beta }_0$ represent constants (intercepts); X is a vector of covariates comprising DT, NR, GI, and PD; $a_i$ and $u_i$ are treated as random effects to capture variations among countries, where each country is treated as a cluster; and $ϵ\mathrm{a}\mathrm{n}\mathrm{d}v$ represent random error terms [90]. To handle the endogeneity problem, biomass energy use was modeled in the first stage using the employment rate as an instrumental variable (Equation (8)), followed by modeling the ecological footprint in the second stage using the predicted values of biomass energy use (${\ln \widehat{\left(BIO\right)}}_{it}$) identified in stage 1 as the primary exposure (Equation (9)).

3.2.3. Local Linear Dummy Variable Estimation Technique

Parametric methods offer estimates of the average impact of biomass energy use on ecological footprints, but lack the ability to depict how this association has changed over time. Consequently, our study aims to employ the nonparametric LLDVE approach to investigate the evolving dynamics of the correlation between biomass energy usage and ecological footprints, as outlined in the following equation.

$$\begin{gathered} {\ln \left(EF\right)}_{it}=f_i\left(t\right)+{\delta }_1\left(t\right)\ln {\left(BIO\right)}_{it}+{\delta }_2\left(t\right){\ln \left(\mathrm{D}\mathrm{T}\right)}_{\mathrm{i}\mathrm{t}}+{\delta }_3\left(t\right){\ln \left(NR\right)}_{it}+ \\ {\delta }_4\left(t\right){\ln \left(GI\right)}_{it}+{\delta }_5\left(t\right){\ln \left(PD\right)}_{it}+u_i+{ϵ}_{it} \end{gathered}$$

(10)

$f_i\left(t\right)=f_i(t/T)$, where, $i=1,2,\dots .,N,$ represents country-specific trend functions; $t=1,2,\dots .,N,$ specifies the time period. Unknown time-varying coefficients are represented by ${\delta }_j\left(t\right)={\delta }_j(t/T)$, where, $j=1,2,\dots ,N;$ $u_i$ indicates the unknown country-specific effect; and ${ϵ}_{it}$ signifies stationery for every $i$. Moreover, it is expected that $\sum _{i=1}^N{u}_i=0$ and $f\left(t/T\right)=f_i\left(t/T\right)$ $\forall$ $i$ (common trend function), and the nation’s trend, can be measured using the errors following Phillips [91]. A detailed picture of the LLDVE method is provided in Awaworyi Churchill et al. [92,93], Hailemariam et al. [94], Silvapulle et al. [95], and Ivanovski et al. [96].

4. Results and Discussion

The results and discussion unit are segmented into three parts. Firstly, in Section 4.1, we concisely summarize the findings from the cross-sectional dependency and slope heterogeneity tests. Subsequently, Section 4.2 elaborates on the parametric estimates obtained through the generalized two-stage least squares (G2SLS) estimation technique. Lastly, Section 4.3 presents the nonparametric estimates derived from the LLDVE approach.

4.1. Results of Cross-Sectional Dependency and Slope Heterogeneity

Table 4. Test for cross-sectional dependence.

Variables	LM		CD
	Test Value	p-Value	Test Value	p-Value
lnEF	3226.89	<0.001	23.03	<0.001
lnBIO	2456.82	<0.001	8.77	<0.001
lnNR	3659.62	<0.001	33.60	<0.001
lnGI	12,742.83	<0.001	112.78	<0.001
lnDT	13,589.45	<0.001	116.56	<0.001
lnPD	10,247.43	<0.001	53.57	<0.001

presents detailed findings of cross-country dependency using both the LM and CD approaches.

The obtained p-values from the LM and CD tests indicate rejection of the null hypothesis concerning cross-country independence for variables such as ecological footprint, biomass energy use, natural resources, globalization index, digitalization, and population density. Hence, all variables in this research exhibit cross-sectional dependence. To explore slope heterogeneity in the data, we utilized the technique proposed by Pesaran and Yamagata [87] with detailed findings displayed in

Table 5. Test for slope homogeneity.

Tests	LM Value	p-Value
$\tilde{\mathrm{\Delta }}$	2.347 **	0.019
${\tilde{\mathrm{\Delta }}}_{adjusted}$	2.745 ***	0.005

*** p-value < 0.01, ** p-value < 0.05.

.

The existence of slope heterogeneity is supported by the pertinent p-values presented in Table 5.

4.2. Parametric Results

To assess the effect of biomass energy usage on ecological footprints while accounting for cross-sectional dependency, we utilized a random effects linear model within a panel framework. Control variables such as natural resources, globalization, population density, and digitalization were incorporated. Detailed results are displayed in

Table 6. Random effects model estimates (baseline results).

Variables	Coefficients (SE)	95% CI
lnBIO	0.376 *** (0.036)	(0.306–0.447)
lnNR	−0.006 (0.006)	(−0.017–0.005)
lnGI	0.500 *** (0.073)	(0.357–0.643)
lnPD	−0.052 * (0.029)	(−0.108–0.005)
lnDT	−0.011 ** (0.006)	(−0.022–0.000)
Random effects parameters
Cluster variance (${\sigma }_u^2$)	0.212
Error variance (${\sigma }_e^2$)	0.096
ICC ($\rho$)	0.829
Breusch and Pagan Lagrangian multiplier test for random effects (${\chi }^2$)	7425.30 ***

Note. *** p < 0.01, ** p < 0.05, * p < 0.1.

.

The estimations of intra-class correlation (ICC) reveal significant variation in the ecological footprint across nations, prompting its inclusion in the random effects linear model to ensure robust conclusions regarding the study objectives. Accounting for country-level variations, the results demonstrate a statistically significant increase in the ecological footprint associated with biomass energy usage across all selected nations. For instance, 1 percent rise in biomass energy use corresponds to a 0.376 percent upsurge in the average ecological footprint.

Nevertheless, when endogeneity is present, the random effect model might produce estimates that are biased. The findings from the Hausman test (refer to

Table 7. Random effects estimates using generalized two-stage least squares.

2nd Stage (lnEF)
Variables	Coefficients (SE)	95% CI
lnBIO	1.824 *** (0.366)	(1.106–2.541)
lnNR	0.023 ** (0.011)	(0.003–0.044)
lnGI	−0.442 * (0.241)	(−0.915–0.030)
lnPD	0.270 ** (0.137)	(0.002–0.537)
lnDT	0.015 (0.010)	(−0.005–0.035)
Hausman test for endogeneity (${\chi }^2$)	2015.39 ***
First-Stage (lnBIO)
InER	0.324 *** (0.065)	(0.196–0.452)

Note. *** p < 0.01, ** p < 0.05, * p < 0.1.

) reveal a notable disparity in the regression estimates between models incorporating and excluding endogeneity, thereby confirming the endogenous nature of biomass energy consumption.

Thus, this study utilized a generalized two-stage least squares random effects regression to address concerns of endogeneity, employing the employment rate as an instrumental variable. The first stage analysis discloses a significant connection between employment rates and biomass energy consumption in OECD nations, which aligns with expectations. As depicted in Table 7, the instrumented estimates (1.824) surpass the baseline results (0.376), indicating that endogeneity in biomass energy usage leads to a downward bias in our baseline estimates. Consequently, it can be inferred that biomass energy use substantially elevates ecological footprints in the studied nations. Furthermore, the G2SLS results suggest that natural resources and population density amplify ecological footprints, while globalization diminishes them. These findings are consistent with prior research conducted in G7 countries [22].

4.3. Nonparametric Results

Point-based estimates offer insights into the average association between variables, yet they do not elucidate the evolving connection between ecological footprint and biomass energy consumption over time. To address this gap, common trends and time-varying nonparametric results for the panel are depicted in Figure 1. This figure showcases the estimates for the time-varying coefficient functions and the common trends alongside their corresponding 95 percent confidence intervals.

Significant time-varying coefficients are interpreted based on their position relative to the zero-axis. Hence, if any of the 95 percent confidence intervals persist (or vary) along the zero-axis, it is assumed that there are no significant time-varying coefficients. The figure indicates a nonlinear common trend function for ecological footprints, showing a slight increase up to 2003, followed by a decline from 2003 to 2010, and then a subsequent rise. Notably, the link between biomass energy consumption and ecological footprints is observed to be time-varying. Over the period, the coefficient exhibits a nonlinear trend, consistently positive and statistically significant for most of the study period, except from 2009 to 2012, when it becomes nonsignificant. These findings validate the conclusions drawn from the parametric estimation, indicating that biomass energy consumption increases ecological footprints across all time periods.

This outcome supports assertions and findings from extensive research, suggesting that biomass energy usage adversely affects environmental quality. Nonetheless, enhancing the efficiency of biomass energy production could mitigate costs, promoting its use as a substitute for fossil fuels. Therefore, a substantial rise in biomass energy production could alleviate dependence on fossil fuels while addressing environmental concerns [71]. However, it is also crucial to acknowledge that food crops and hydrocarbon-rich plants primarily serve as biomass sources for energy generation [96]. While expanding these sources may absorb carbon emissions, the associated environmental impacts of biomass energy extraction in OECD nations cannot be overlooked. Issues such as soil degradation, nutrient and water depletion, and deforestation may arise from energy crop cultivation. Additionally, biomass harvesting and combustion can pose further environmental risks. While debates surrounding the environmental consequences of biomass energy use in OECD countries persist [21,68,72], these empirical findings provide a foundation for policymakers to understand the negative consequences of biomass energy use.

To mitigate the ecological footprint of biomass energy and promote its sustainable use, policymakers should implement a multifaceted strategy. This includes enforcing certification standards for sustainable biomass sourcing, ensuring that feedstock materials are derived from responsibly managed crops, forest residues, and waste. Investment in advanced biomass technologies should be prioritized to enhance efficiency and reduce emissions. Additionally, diversifying renewable energy sources will further minimize reliance on biomass, spreading the ecological impact across various sustainable options. Policymakers must also establish rigorous monitoring systems to track and address the environmental risks associated with biomass production. Public education campaigns are essential to raise awareness about the benefits and practices of sustainable energy, fostering greater community support and understanding. These measures, supported by robust regulatory frameworks, will help align biomass energy consumption with global climate goals, and reduce its ecological footprint.

5. Conclusions

This study investigated the impact of biomass energy consumption on ecological footprints across OECD nations from 1990 to 2017 while considering factors such as natural resources, globalization, digitalization, and population density. The traditional literature predominantly relies on parametric approaches, offering estimates for the average effect of biomass energy use on ecological footprints. However, these methods fail to capture how this relationship evolves over time. To address this gap, we employ both parametric (G2SLS) and nonparametric time-varying LLDVE techniques to assess the impact of biomass energy use on environmental footprints. Results from both parametric and nonparametric methods indicate that biomass energy use correlates with higher ecological footprints. For instance, the generalized two-stage least square method suggests that a 1 percent rise in biomass energy usage corresponds to a 1.824 percent upsurge in the average ecological footprint. Conversely, the LLDVE method reveals a time-varying relationship between biomass energy consumption and ecological footprints. It demonstrates predominantly positive and statistically significant estimated coefficients throughout the analyzed time period, indicating a consistent increase in ecological footprints due to biomass energy use across the studied nations.

Despite these contributions, this research has limitations that warrant attention in future studies. Due to data constraints, environmental drawbacks associated with consuming different forms of biomass energy were not considered. Moreover, our analysis focused solely on OECD countries, which are predominantly developed nations. Future research could expand on our findings by exploring the impact of biomass energy use on the environment across a wider range of countries and economic development levels, incorporating additional data and variables to understand the holistic mechanism between ecological footprint and biomass energy use.