Socio-Economic Drivers of Renewable Energy Consumption: A Dynamic Panel Analysis of Rural and Urban Contexts in Europe

Abstract

This study examines the patterns of renewable energy consumption across 29 European countries from 2000 to 2024. We utilised Bias-Corrected estimation techniques to analyse the relationship between renewable energy consumption, Human Development Index (HDI), labour force, and population dynamics, employing three distinct estimation models: global, rural, and urban. The results indicate that higher levels of human development and a larger labour force are positively associated with renewable energy consumption in the global and rural models, while the urban model shows an opposite effect for the labour force. Conversely, population growth is negatively related to renewable energy consumption in the global and rural contexts but positively in urban areas. These findings underscore the importance of socio-economic and demographic contexts in shaping renewable energy outcomes. They suggest that renewable energy can support economic and social development, but its effectiveness depends on regional structural conditions. From a policy perspective, the renewable transition should be approached as both an environmental and socio-economic strategy, aligning clean energy goals with employment generation, education, and inclusion, particularly in line with SDGs 7, 8, 10, and 13. Policies that promote green skills, innovation, and equitable regional investment can enhance social acceptance, competitiveness, and sustainable growth across Europe.

1. Introduction

Energy lies at the heart of the planet’s ecological and socio-economic systems. It not only drives biological processes but also underpins industrial activities, urban development, and regional livelihoods. However, the planet is experiencing a paradigm shift because there is a dependence on fossils, which is not sustainable. The consumption of fossil fuels has contributed to the increase in greenhouse gas (GHG) emissions, exacerbating the threat of climate change and global warming.

The promotion of renewable energy has become a strategic priority in the global policy agenda, particularly in the context of achieving the targets set by the Paris Agreement and the United Nations Sustainable Development Goals (SDGs), especially SDG 7, which emphasises affordable and clean energy for all [1]. The European Union (EU) has emerged as a global leader in this transition, with member states collectively increasing their share of renewables in gross final energy consumption from 9.6% in 2004 to 22.5% in 2022 [2]. However, despite this progress, spatial disparities in renewable energy consumption persist across and within countries, particularly between rural and urban areas.

Socio-economic development plays a key role in shaping a country’s energy mix, particularly the shift toward renewable sources. Economic growth fosters the infrastructure and investments necessary for the adoption of renewable technologies, and, in turn, access to sustainable energy enhances economic productivity, education, and public health. Nonetheless, the drivers of renewable energy consumption may differ substantially between rural and urban settings due to structural, geographic, demographic, and policy-related factors. For example, rural areas may benefit from greater land availability and proximity to biomass or hydropower resources, whereas urban regions may face regulatory or space constraints but enjoy better technological and financial infrastructure for solar or district energy systems (district energy systems are centralised infrastructures that produce and distribute thermal energy (heating or cooling) to multiple buildings within a neighbourhood or district through an underground network of pipes [3]) in [4]. The relationship between renewable energy and human development in general is beneficial, as pointed out by [5,6,7,8] (human development consists of expanding people’s capabilities and freedoms, broadening their set of choices, and ensuring that all citizens enjoy their human rights; it is a conception of development that goes beyond mere economic growth and places people’s lives and well-being at the centre [9]). The relationship tends to be even more beneficial in countries with a high level of education [7,8].

In addition, the transition to the renewable energy matrix can contribute to job creation, as it has been shown to reduce unemployment rates by creating jobs directly in the construction and maintenance of renewable energy infrastructure and indirectly through increased productive efficiency and better conditions for business [10,11,12]. Also, the development of renewable energy can positively affect the relative population of rural regions by creating local employment opportunities and improving living conditions, which can help retain and attract residents [10,12]. However, most existing studies analyse these relationships at the national level, overlooking the heterogeneity of drivers between rural and urban areas, and rarely integrate socio-economic, demographic, and labour market variables into a unified analytical framework. This limits our understanding of how renewable energy interacts with human development and employment across distinct territorial contexts.

From a theoretical perspective, this study is grounded in the endogenous growth framework, which highlights the role of human capital, technological diffusion, and institutional capacity as engines of sustainable growth. Within this context, renewable energy acts not only as an environmental necessity but also as a driver of productivity and regional development. Theoretically, differences in regional capabilities and socio-economic structures imply that the determinants of renewable energy consumption may vary systematically across rural and urban settings.

In light of this context, the present study seeks to explore the differential determinants of renewable energy consumption across rural and urban areas in 29 European countries from 2000 to 2024. Specifically, this research addresses the following questions:

• What are the key drivers of renewable energy consumption in rural versus urban contexts?
• How do these determinants vary across space, and what are the policy implications for fostering an inclusive energy transition?

The specific objectives are threefold: (i) to identify the key socio-economic, demographic, and institutional drivers of renewable energy consumption in rural and urban regions; (ii) to assess how these determinants differ across space and over time; and (iii) to derive policy implications for promoting an inclusive and balanced renewable energy transition.

To the best of our knowledge, previous empirical research has not distinctly examined these regional dynamics within the European context using a unified analytical framework. This study contributes to filling this gap by employing panel-data techniques to assess the factors influencing renewable energy uptake across differing spatial domains. Section 2 concisely reviews the relevant literature on the subject, while Section 3 provides details on the data and approach used in the investigation. Section 4 presents the study’s results, and Section 5 presents the discussion and analysis based on the research objectives. Finally, Section 6 concludes with some closing remarks, limitations of this study, and suggestions for future research.

2. Literature Review

As is widely acknowledged, economic and social development typically represents a primary goal for governing authorities (According to the United Nations, social development is an inclusive concept that refers both to the well-being of individuals and to the harmonious functioning of societies. It involves expanding economic opportunities, reducing poverty, ensuring access to employment and public services, and promoting social progress and justice so that everyone has a real voice and stake in their community. Social development integrates economic growth with social welfare and environmental protection, aiming for an inclusive and just world where no one is left behind [13]). This pursuit is driven by the expectation that improving the population’s quality of life will, in turn, increase the likelihood of political leaders remaining in office. In this section, we undertake a concise review of the contemporary literature about social and economic development over the recent years.

2.1. Economic Growth and Renewable Energy Consumption

In [14], a study encompassing 34 economies, positive impacts of renewable energy consumption on economic growth were observed in 22 economies, while negative effects were identified in three countries, with no significant results found in nine countries. Ref. [15] identified a bidirectional causal relationship between GDP and renewable energy consumption in G7 countries. In a study by [16], which analysed 68 developing economies, positive relationships were found between CO₂ emissions, renewable and non-renewable energy consumption, and economic growth. Ref. [17] presents an intriguing perspective suggesting that the substitution of fossil fuels with renewable energy sources could potentially have adverse effects on human development, despite its positive environmental impact. To investigate this hypothesis, the researchers examined the relationship between renewable energy and human development across 28 OECD (Organisation for Economic Cooperation and Development) countries spanning from 1990 to 2017. Contrary to [15]’s initial expectations, the data analysis revealed a positive impact of renewable energy on human development. This finding challenges the initial belief and indicates that the adoption of renewable energy sources can indeed contribute positively to human development. Furthermore, the causality test conducted by the researchers indicated a bidirectional causality relationship between renewable energy and human development, highlighting the complex interplay between these variables [17]. However, in order to sustain the complex relationship between these variables, the study also found a unidirectional relationship from the HDI to renewable energy consumption among the 10 countries with the highest HDI [5].

Ref. [18]’s proposed study delves into the reciprocal relationship between human development and sustainable development. As posited by the authors, enhancements in the HDI correlate with reductions in carbon dioxide, GHG, and ecological footprint emissions [18]. In the context of the Chinese economy, ref. [19] reveals a negative correlation between the HDI and CO₂ emissions. This suggests that progress in human development can potentially mitigate the effects of pollution, indicating a promising avenue for environmental sustainability [19], a result that is sustained for Pakistan. According to [6], there is a positive association between economic growth and CO₂ emissions; however, it was also observed that human development, technology, and renewable energy boost economic development. In the same vein, empirical evidence indicates that access to clean energy, per capita income, and technical progress contribute to an improvement in the HDI; on the other hand, urbanisation is a factor that tends to impair the quality of life of individuals [8]. The relationship between renewable energy, information and communication technologies (ICTs), and human development should be investigated. One such analysis, covering 26 countries from 2000 to 2018 through static and dynamic panel models, offers nuanced insights into these relationships [20]. The results indicate that renewable energy alone does not have a statistically significant impact on the HDI in either the short or long term. Conversely, ICTs show a positive effect on HDI in the short term, although this effect dissipates over time. Notably, the combination of renewables and ICTs yields a significant positive influence on HDI across both time horizons, suggesting potential synergies between technological advancement and clean energy in promoting human development. Additionally, economic development consistently enhances HDI, while carbon dioxide emissions exert a negative influence. Primary energy consumption appears to support HDI growth in the short term but becomes detrimental over longer periods. Lastly, long-term population growth is associated with declines in HDI [20].

In China, the shift towards renewable energy and urbanisation is expected to reduce the rural population, potentially modernising agriculture and increasing rural incomes [21]. In a broader context, it is plausible to infer that human development enhances overall well-being and fosters environmental amelioration. However, a contrasting relationship was noted in the specific case of Latin America, where an upswing in human development appears to be linked with a considerable environmental toll [22]. Ref. [23] suggests that urbanisation, biomass consumption, economic growth, and globalisation collectively tend to have a positive influence on human development in this context [23].

2.2. Human Development and Renewable Energy

Ref. [17] posits an intriguing perspective, suggesting that the replacement of fossil fuels by renewables could potentially hurt human development, despite its positive environmental impact. To investigate this hypothesis, researchers examined the relationship between renewable energy and human development across 28 OECD countries from 1990 to 2017. Contrary to [17] expectations, the data analysis revealed a positive impact of renewable energy on human development. This outcome challenges the initial belief and suggests that adopting renewable energy sources can contribute positively to human development. Additionally, the causality test indicated a bidirectional causality relationship between renewable energy and human development, further emphasising the complex interplay between these variables [17]. In European countries, renewable energy production significantly reduces unemployment in the long run, with positive changes in renewable energy production leading to a decrease in unemployment [11]; a similar result was observed in Ecuador [10].

In this context, the research undertaken by [24] endeavours to examine the interrelationship between Greenhouse Gas (GHG) emissions and the Human Development Index within the context of eight newly admitted members of the European Union. The findings derived from this investigation reveal a discernible negative correlation between these variables [24]. In a broader context, it is plausible to infer that human development enhances overall well-being and fosters environmental amelioration. However, a contrasting relationship was noted in the specific case of Latin America, where an upswing in human development appears to be linked with a considerable environmental toll [22].

Ref. [25] extensive study examined the intricate relationship between renewable energy consumption, CO₂ emissions, and the HDI across a broad spectrum of 126 economies. While the findings exhibit a degree of heterogeneity, a prevailing pattern emerges across the majority of nations under investigation. Specifically, the results indicate a positive influence of renewable energy consumption on HDI, signifying that greater utilisation of renewable energy sources tends to enhance human development. However, it is noteworthy that exceptions to this trend are evident in the MENA (Middle East and North Africa) and Central America and Caribbean regions, where the consumption of renewables appears to harm human development. Conversely, in Europe, the effect is positive, indicating a favourable synergy between renewable energy use and human development. Remarkably, a contrasting relationship is observed concerning carbon dioxide emissions, which exhibit a positive correlation with HDI across all regions [25]. This outcome suggests that the increase in CO₂ emissions is intertwined with economic growth, and there appears to be a direct association between emissions and economic development. When oil products, natural gas, and total energy consumption were employed as indicators of energy use, the findings demonstrated a negative effect on human development in the short term, with no statistically significant influence observed over the long term. In contrast, using electricity, coal, and lignite as energy proxies yielded no significant effects, regardless of the time period analysed. Overall, the results suggest that while renewable energy contributes positively to human development, its benefits are insufficient to counterbalance the adverse effects associated with other, more traditional energy sources [26].

2.3. Regional Contexts, Energy Consumption and Human Development

In rural European regions, renewable energy contributes positively to local income growth, offering an avenue for socio-economic revitalisation. However, this optimistic view is tempered by evidence indicating that in areas facing population decline, the deployment of renewable energies may not effectively counteract depopulation trends and, in some cases, may even exacerbate them [12]. A study analysing panel data from 104 countries between 2000 and 2020 explored the relationship between globalisation, renewable energy use, labour force participation, and sustainable development [27]. The findings highlight that all three factors—globalisation, labour availability, and renewable energy consumption—exert a positive influence on sustainable development outcomes. Notably, the study emphasises that increased renewable energy use consistently enhances sustainability across different subgroups of countries, reinforcing its central role in advancing long-term development goals [27].

A study analysed the relationship between renewable and non-renewable energy consumption, carbon dioxide emissions, urbanisation, industrialisation and the HDI for eight economies considered to be developing. The findings point to a U-shaped relationship between renewable energy consumption and HDI and an inverted U-shaped relationship between HDI and fossil energy/CO₂ emissions [28]. That is, the consumption of renewable energies initially tends to harm the quality of life, but this relationship is subsequently reversed, while the opposite is observed for non-renewable energies. Finally, the positive impact of urbanisation, international trade and industrialisation on HDI [28]. In the same sense, ref. [29] examines the asymmetric role of renewable energy, carbon emissions, economic growth, and urbanisation on human development in European Union countries, the results corroborate what has been previously observed, renewable energies and HDI have a beneficial relationship, as well as urbanisation and economic growth [29].

The existing literature consistently shows that renewable energy consumption is shaped by economic growth, human development, technological progress, and demographic dynamics, though the magnitude and direction of these effects vary across countries and development levels. Studies also highlight that urbanisation and employment structures influence energy demand, while regional disparities create heterogeneous pathways for the adoption of renewable technologies. Despite these important insights, prior research has overwhelmingly analysed countries as homogeneous units, overlooking how the determinants of renewable energy consumption differ between rural and urban contexts, where infrastructure availability, income levels, population density, and policy incentives operate in distinct ways. Consequently, the literature lacks a comprehensive cross-country assessment that explicitly compares the drivers of renewable energy consumption in rural and urban areas over time. By addressing this gap, this study contributes new empirical evidence on the spatial heterogeneity of renewable energy determinants across Europe, offering a more nuanced understanding of how socio-economic and demographic factors shape the transition toward sustainable energy systems.

3. Data and Methods

3.1. Data

The research data utilised in this study were sourced from secondary online databases. Specifically, the dependent variable, focusing on the consumption of renewable energies, was obtained from the 2025 Energy Institute Statistical Review of World Energy database. Meanwhile, the explanatory variables encompassing rural and urban population statistics, as well as employment data in rural and urban areas, were extracted from the World Development Indicators (WDI). Additionally, information about the Human Development Index (HDI) was gathered from the United Nations Development Programme (UNDP). The division between the rural and urban models was calculated on a per capita basis. The WDI database provides data on the total population and the percentage distribution of individuals living in urban and rural areas, which were used for this purpose. The survey’s sample period spans from 2000 to 2024, encompassing data from 29 European countries. These countries include Austria, Belgium, Croatia, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, the Netherlands, Norway, Poland, Portugal, Slovakia, Slovenia, Spain, Sweden, Switzerland, Turkey, Ukraine, and the United Kingdom. The table below presents a summary of the model variables.

The three estimation models, global, urban, and rural, were constructed using data obtained from secondary sources (see

Table 1. Variables.

Variables	Unit Measure	Source
Renewables Consumption	Exajoules	2025 Energy Institute Statistical Review of World Energy
Human Development Index (HDI)	Index	United Nations Development Programme (UNDP)
Population, total	Total	World Development Indicators (WDI)
Labour Force, total	Total	World Development Indicators (WDI)
Rural population, total	Total	World Development Indicators (WDI)
Rural Labour, total	Total	World Development Indicators (WDI)
Urban Population, total	Total	World Development Indicators (WDI)
Urban Labour, total	Total	World Development Indicators (WDI)

). The global model encompasses the total population of the countries analysed, integrating all available national-level indicators. In contrast, the urban and rural models were developed to capture the heterogeneity across different population contexts. For these models, data from the World Development Indicators (WDI) were used to obtain the percentage of the population residing in urban and rural areas. By combining these percentages with the total population, it was possible to estimate the number of inhabitants in each region, allowing the construction of separate models that reflect specific regional dynamics. This approach enables comparative analysis between global, urban, and rural dimensions, highlighting potential differences in the determinants of the studied phenomenon across distinct population distributions.

Table 2. Descriptive Statistics.

	Renewables Consumption	HDI	Population, Total	Labour Force, Total	Rural Population	Rural Labour	Urban Population	Urban Labour
Mean	0.171495108	0.879079054	21,035,726.1	9,934,860.482	5,370,249.4	251,111,002	15,665,477	5,479,463,830
Median	0.052334324	0.889	9,519,374	4,646,803	2,209,932	115,248,095	6,546,012	329,267,998
Standard deviation	0.31545374	0.054990197	25,179,855.5	11,689,786.12	6,509,755.1	292,370,696	19,101,347	7.3531 × 1010
Minimum	1.14919 × 10−5	0.67	281,205	169,444	20,437	1,202,758.2	259,836	15,656,795
Maximum	2.436822948	0.972	85,518,661	44,433,744	22,605,796	1.005 × 109	66,613,051	1.1479 × 1012
Observations	725	725	725	725	725	725	725	725

below presents the descriptive statistics of the variables used to estimate the model.

3.2. The Bias-Corrected Estimator (BC)

In the analysis of models that include lagged variables and unobserved heterogeneity, models with fixed and random effects can produce biased results in contexts in which the time horizon is short. In other words, the estimators obtained from these approaches may not accurately reflect the observed relationships between the variables, so the estimators must be chosen carefully.

Considering that the bias of the estimators is known analytically, the bias-corrected (BC) estimator is able to eliminate it directly at its source by adjusting the momentum conditions. In addition, fixed effects (FE) and random effects (RE) estimators retain their low-variance properties. The BC estimator is compatible with higher-order autoregressive models and can be implemented in versions that incorporate FE or RE. It is a time-based method, with a well-defined asymptotic distribution, which facilitates the calculation of standard errors. These errors can be further adjusted to account for the dependence between cross-sections, using robust techniques [30].

A generic Dynamic panel-data model can be described as:

$$y_{it}=\sum _{j=1}^p\lambda y_{i,t-1}+x_{it}^{\prime }\beta +{\epsilon }_{it}\hspace{1em}\hspace{1em}{\epsilon }_{it}=u_i+{\alpha }_{it}$$

(1)

In which $x_{it}^{\prime }$ is a vector of time-varying variables [31]. In dealing with panel-data estimation, it is necessary to address whether the fixed or random effects will be applied; this decision is taken through the interpretation of Hausman’s Test.

The main assumptions of the model are as follows: it is a high-order autoregressive model in which, under minimal conditions of regularity in the initial observations, the dependent variable can be explained by up to $p$ lags. The regressors, represented by $x_{it}$, are considered strictly exogenous, implying that $E\left[x_{it}{u}_{is}\right]=0$ for all periods $t$ and $s$, relative to the idiosyncratic error term. The model contemplates the presence of specific effects that are not observable by group, such as fixed effects (FE), if $E\left[x_{it}{\alpha }_i\right]\ne 0$, or for RE $E\left[x_{it}{\alpha }_i\right]=0$. It is also assumed that idiosyncratic errors do not present serial autocorrelation, i.e., the expected value of the product between errors in different periods is null. Even so, the possibility of heteroscedasticity is admitted, and the variance of errors can vary between units, denoted by ${\sigma }_i$ [30].

To simplify the model, it is assumed that $p=1$ and $\theta ={\left({\lambda }_1,{\beta }^{\prime }\right)}^{\prime }$. The estimator, known as the fixed effects and bias-corrected estimator with just identification, solves the following:

$$\widehat{\theta }=arg\hspace{0.17em}\underset{\theta }{\min }{\left(\sum _{i=1}^N{m}_i\left(\theta \right)\right)}^{\prime }\left(\sum _{i=1}^N{m}_i\left(\theta \right)\right)$$

(2)

While the estimator, referred to as the random effects and bias-corrected estimator with over-identification, solves the following:

$${\widehat{\theta }}^{\left(j\right)}=arg\hspace{0.17em}\underset{\theta }{\min }{\left(\sum _{i=1}^N{m}_i\left(\theta \right)\right)}^{\prime }W\left(\sum _{i=1}^N{m}_i\left(\theta \right)\right)$$

(3)

In the context of finite element analysis, we used an adjusted profile likelihood estimator as the basis for estimating the bias-corrected model (BC). This estimator was developed specifically for contexts in which the dependent variable has only one lag. In addition to the advantages already highlighted, it is important to note that this estimation approach does not require the prior use of a consistent estimator, which reinforces its applicability and efficiency [30].

The inclusion of time effects improves the model’s robustness by capturing common temporal shocks and mitigating bias due to omitted variables that vary over time but not across entities [32]. In contrast, models that omit time effects may suffer from omitted variable bias, particularly when time-varying shocks are correlated with explanatory variables [33]. The BC estimator, initially proposed by [34] and refined for small panels by [35], adjusts the finite-sample bias of the least-squares dummy variable estimator (LSDV) and retains valid standard errors even when the time dimension is limited. When estimating the model without time effects, the BC estimator corrects primarily for the bias induced by the inclusion of a lagged dependent variable and unit-specific fixed effects [35]. In contrast, bias-corrected estimation with time effects additionally adjusts for the degrees of freedom consumed by time dummies and more precisely isolates the net impact of the regressors from common shocks [32]. This dual approach allows us to assess the sensitivity of our findings and ensures that the results are not artefacts of time-related confounding.

4. Results

Three separate models were estimated in this study to analyse the determinants of the outcome variable. The first model, termed the “global model,” encompasses data from the entire country without any spatial subdivision. The second model, referred to as the “rural model,” focuses solely on observations from rural areas, excluding urban areas. Conversely, the third model, denoted as the “urban model,” restricts the analysis to urban areas, excluding rural regions. By employing these distinct models, the study aims to examine potential variations in the determinants of the outcome variable across different geographical contexts, thereby providing insights into the heterogeneous nature of the relationships under investigation.

Table 3. Bias-Corrected Estimations Global Model.

Global Model	Random Effects	Fixed Effects	Global Model	Random Effects	Fixed Effects
Renewables Consumption			Renewables Consumption
L1	0.9569628 ***	0.9189123 ***	L1	1.119347 ***	1.043958 ***
			L2	−0.1835345 ***	−0.137648 **
HDI	0.1216745 ***	0.0208206	HDI	0.1363215 ***	0.0395965
Population, total	−5.29 × 10−10	−4.40 × 10−9	Population, total	−1.10 × 10−9 *	−3.99 × 10−9
Labour force, total	2.83 × 10−9 *	1.82 × 10−8 **	Labour force, total	4.32 × 10−9 ***	1.76 × 10−8 **
Constant	−0.1052551 ***	−0.0819797	Constant	−0.1192051 ***	−0.1003557
Hausman Test	30.6681 ***		Hausman Test	35.8389 ***
Hansen Test	6.2037 ***		Hansen Test	5.1725 **
Arellano–Bond test for Autocorrelation			Arellano–Bond test for Autocorrelation
H0: no autocorrelation of order 1:	−2.1161 **		H0: no autocorrelation of order 1:	−2.2481 **
H0: no autocorrelation of order 2:	−1.0777		H0: no autocorrelation of order 2:	0.6572

Note: significance level at 1% ***, 5% **, and 10% * to evaluate the results.

,

Table 4. Bias-Corrected Estimations Rural Model.

Rural Model	Random Effects	Fixed Effects	Rural Model	Random Effects	Fixed Effects
Renewables Consumption			Renewables Consumption
L1	0.9669648 ***	0.9425973 ***	L1	1.126454 ***	1.062669 ***
			L2	−0.1720791 ***	−0.1350507 **
HDI	0.1557724 ***	0.0576697	HDI	0.1626763 ***	0.0830489
Rural population	−2.04 × 10−9	−2.68 × 10−8	Rural population	−3.60 × 10−9	−2.63 × 10−8
Rural Labour	1.05 × 10−10	5.76 × 10−10 ***	Rural Labour	1.44 × 10−10 **	5.04 × 10−10 ***
Constant	−0.136089 ***	−0.0303028	Constant	−0.1430428 ***	−0.0365815
Hausman Test	2.6607 *		Hausman Test	4.4496 ***
Hansen Test	6.8072 ***		Hansen Test	5.6588 ***
Arellano–Bond test for Autocorrelation			Arellano–Bond test for Autocorrelation
H0: no autocorrelation of order 1:	−2.1186 **		H0: no autocorrelation of order 1:	−2.1748 **
H0: no autocorrelation of order 2:	−1.0837		H0: no autocorrelation of order 2:	0.5274

Note: significance level at 1% ***, 5% **, and 10% * to evaluate the results.

and

Table 5. Bias-Corrected Estimations Urban Model.

Urban Model	Random Effects	Fixed Effects	Urban Model	Random Effects	Fixed Effects
Renewables Consumption			Renewables Consumption
L1	0.9817621 ***	0.9373087 ***	L1	1.131017 ***	1.060418 ***
			L2	−0.1600688 ***	−0.1394077 **
HDI	0.0891199 **	0.0767583	HDI	0.090663	0.0974547 *
Urban population	6.53 × 10−10 **	4.94 × 10−9 ***	Urban population	6.86 × 10−10 **	5.25 × 10−9 ***
Urban Labour	−5.05 × 10−14 ***	−5.03 × 10−14 ***	Urban Labour	−4.30 × 10−14 ***	−4.39 × 10−14 ***
Constant	−0.0749166 *	−0.1230456 ***	Constant	−0.0767198	−0.1447108 ***
Hausman Test	15.7072 ***		Hausman Test	8.8637 ***
Hansen Test	4.4788 **		Hansen Test	4.5356 **
Arellano–Bond test for Autocorrelation			Arellano–Bond test for Autocorrelation
H0: no autocorrelation of order 1:	−2.1549 **		H0: no autocorrelation of order 1:	−2.0914 **
H0: no autocorrelation of order 2:	−1.1105		H0: no autocorrelation of order 2:	0.3933

Note: significance level at 1% ***, 5% **, and 10% * to evaluate the results.

present the estimation results for the three proposed models, each estimated using both fixed and random effects methodologies and with a lag. To determine the most appropriate estimator for each model, the Hausman test was employed. The Hausman test assesses whether the differences between the estimates obtained from the fixed effects and random effects models are systematic or random. For the models without lag, the null hypothesis is consistently rejected across all models, thus indicating that the fixed effects model stands as the most suitable approach for analysis. Conversely, when lag is introduced, the null hypothesis remains unchallenged across all models. Consequently, no clear distinction emerges among the models, rendering each equally viable for analysis. As a result, for models incorporating lag, the decision was made to employ those featuring random effects, owing to their demonstrated robustness in yielding results.

In the specification without lag, the HDI remains statistically significant across all models—global, rural, and urban (see Table 3, Table 4 and Table 5)—although its magnitude and level of significance are more pronounced in the global and rural frameworks. When lagged effects are introduced, the global model reveals that all explanatory variables are statistically significant, indicating that each exerts a meaningful influence on renewable energy consumption. The positive coefficients associated with HDI and the labour force confirm that higher levels of human development and greater labour participation contribute to increased renewable energy consumption. In contrast, the population variable displays a negative relationship, suggesting that population growth, when controlling for other factors, is associated with a relative decline in renewable energy consumption. This pattern is largely consistent across the Global and Rural models, where HDI and labour continue to exert positive effects—particularly in rural contexts—while the population variable demonstrates weaker or negative associations. In the urban model, the results shift considerably: population growth becomes a positive driver of renewable energy consumption. This outcome may reflect the higher average income typically found in urban areas, which, combined with targeted incentives for adopting renewable technologies, such as subsidies for installing solar panels, facilitates greater renewable energy production and, consequently, higher consumption. Conversely, labour force expansion is associated with a reduction in renewable energy consumption. Since increased employment often stimulates higher levels of production, it may necessitate a greater reliance on energy sources that are more cost-effective or offer higher output than renewable alternatives. As a result, the growing demand for energy in more productive urban settings may lead firms to substitute toward conventional energy sources, thereby reducing the relative consumption of renewables. Overall, the results underscore the robustness of human development and labour dynamics as key determinants of renewable energy consumption, with the intensity and direction of these relationships varying between rural and urban settings.

Table 6. Bias-Corrected Estimations with Time-Effects Global Model.

Global Model	Random Effects	Fixed Effects
Renewables Consumption
L1	1.186279 ***	1.018786 ***
L2	−0.1678625 ***	−0.1448321 **
HDI	0.0735634 **	−0.584581 *
Population, total	9.47 × 10−10	−5.13 × 10−9
Labour force, total	−1.53 × 10−9 ***	2.43 × 10−8 **
Constant	−0.0647444 **	0.3791405
Year Dummies	YES
Hausman Test	1.79 × 104 ***
Arellano–Bond test for Autocorrelation
H0: no autocorrelation of order 1:	−1.7759 *
H0: no autocorrelation of order 2:	0.5573

Note: significance level at 1% ***, 5% **, and 10% * to evaluate the results.

,

Table 7. Bias-Corrected Estimations with Time-Effects Rural Model.

Rural Model	Random Effects	Fixed Effects
Renewables Consumption
L1	1.126529 ***	1.054025 ***
L2	−0.1783077 ***	−0.1341321 **
HDI	0.2158271 ***	−0.0667844
Rural population	−3.62 × 10−9	−3.08 × 10−8
Rural Labour	1.46 × 10−10 **	5.81 × 10−10 ***
Constant	−0.2735988 ***	0.0892972
Year Dummies	YES
Hausman Test	436.9077 ***
Arellano–Bond test for Autocorrelation
H0: no autocorrelation of order 1:	−1.9607 **
H0: no autocorrelation of order 2:	−0.1127

Note: significance level at 1% *** and 5% ** to evaluate the results.

and

Table 8. Bias-Corrected Estimations with Time-Effects Urban Model.

Urban Model	Random Effects	Fixed Effects
Renewables Consumption
L1	1.102641 ***	1.05728 ***
L2	−0.1478617 ***	−0.1450077 **
HDI	0.179724	−0.1168306
Urban population	1.02 × 10−9	5.98 × 10−9 ***
Urban Labour	−3.46 × 10−14 ***	−3.24 × 10−14 **
Constant	−0.0475842 *	0.0195921
Year Dummies	YES
Hausman Test	244.3787
Arellano–Bond test for Autocorrelation
H0: no autocorrelation of order 1:	−2.5495 ***
H0: no autocorrelation of order 2:	0.3758

Note: significance level at 1% ***, 5% **, and 10% * to evaluate the results.

present estimates incorporating temporal effects, allowing for the identification of specific years during which variables significantly influenced renewable energy consumption. The Hausman test results indicate no discernible difference between models; thus, the interpretation will rely on random effects and emphasise the temporal aspect. The bias-corrected fixed effects estimations were adopted as the preferred specification across all models, following the rejection of the null hypothesis in the Ramsey RESET test, which indicated the presence of model misspecification under alternative estimators.

The results reveal heterogeneous temporal dynamics in renewable energy consumption across the global, rural, and urban contexts (see Table 6, Table 7 and Table 8).

In the global model (Table 6), renewable energy consumption was significantly influenced by the explanatory variables during multiple years—including 2004–2021—highlighting the dynamic nature of the global energy transition and the sustained impact of structural and policy factors. The predominance of positive year-specific coefficients suggests a consistent expansion of renewable energy utilisation, reflecting worldwide progress in clean energy technologies and policy incentives.

Conversely, in the rural model (Table 7), none of the year dummies were statistically significant, implying that renewable energy consumption in rural areas remained relatively stable over time and was primarily shaped by structural determinants, such as the persistence of past consumption patterns and labour-related factors, rather than by short-term policy or economic shocks.

In contrast, the urban model exhibited significant year effects in 2004, 2008, 2009, 2010, 2012, 2017, and 2019, with marginal significance observed in several other years, indicating that urban renewable energy consumption responded more dynamically to temporal shocks—likely associated with policy initiatives, technological advancements, and changes in energy demand. The consistent positive and significant first lag across models underscores the persistence and path dependency of renewable energy consumption, while the insignificance of the second lag suggests a gradual adjustment process over time.

5. Discussion

The empirical findings reveal a consistent and positive association between human development and renewable energy consumption across all estimated models, reinforcing the hypothesis that improvements in social and economic well-being contribute to the transition toward cleaner energy sources. This outcome aligns with the evidence reported by [17], who found that renewable energy positively influences human development in OECD countries, challenging the notion that energy transitions necessarily entail social costs. Similarly, Refs. [25,29] documented comparable relationships across broad sets of economies, emphasising the synergistic link between renewable energy deployment and enhancements in the HDI. The positive and significant coefficients for HDI in both global and rural models thus reinforce the multidimensional interdependence between clean energy use, human welfare, and sustainable growth.

The results also highlight meaningful spatial heterogeneity. While renewable energy consumption responds positively to HDI and labour force participation in rural and urban contexts, the magnitude of these effects is stronger in rural areas. This finding suggests that renewable energy development in non-urban regions may play a particularly vital role in improving living standards and reducing economic disparities, consistent with [21], who noted that renewable energy expansion in rural China modernised agriculture and increased rural incomes. Conversely, in urban areas, renewable energy consumption exhibits a more dynamic response to temporal shocks, such as technological changes and policy initiatives, echoing the findings of [20] that link technological progress, clean energy use, and HDI improvements through the interaction between renewable energy and information and communication technologies (ICTs).

The persistence of renewable energy consumption across all models, indicated by the significant first lag, underscores the path-dependent nature of clean energy transitions. Similar to [14,15], who identified long-term bidirectional relationships between economic performance and renewable energy use, this evidence suggests that countries with established renewable infrastructures continue to build upon prior consumption trajectories, reinforcing structural commitment to sustainable energy. Moreover, the observed temporal effects in the global and urban models, particularly during key years of policy and technological advancement, mirror the global expansion of renewable energy markets and the broader diffusion of green innovations highlighted by [27].

The negative or weak relationship between population size and renewable energy consumption in some models indicates potential efficiency challenges associated with demographic expansion. As [8] pointed out, rapid urbanisation can undermine quality of life, even when clean energy access improves, reflecting the complexity of balancing demographic growth and sustainability. In contrast, the positive role of the labour force suggests that human capital remains a central driver of renewable energy diffusion. This pattern resonates with [23], who observed that labour participation and economic activity contribute positively to human development and indirectly to sustainable energy transitions.

The study’s findings also align with the notion that human development contributes to environmental improvement. The positive link between HDI and renewable energy supports the argument advanced by [18,19] that rising human development levels correspond with reductions in carbon emissions and ecological footprints. The divergence between rural and urban dynamics, however, cautions against a uniform interpretation of this relationship. While the rural context appears structurally driven and stable over time, the urban model reveals heightened sensitivity to short-term shocks, possibly reflecting policy interventions or technological diffusion patterns similar to those discussed by [22] in their analysis of environmental trade-offs in Latin America.

Finally, the persistence and asymmetry of renewable energy consumption across spatial contexts lend support to the U-shaped and inverted U-shaped relationships reported by [28]. The initial phases of renewable adoption may be constrained by high investment costs or infrastructure barriers, but over time, these constraints give way to enhanced socio-economic and environmental outcomes, as reflected in the increasingly positive coefficients in the later years of the dataset.

Taken together, these findings corroborate prior empirical evidence that renewable energy consumption, human development, and labour dynamics are mutually reinforcing elements of sustainable growth. However, the spatial and temporal heterogeneity observed here underscores that policy design should be context-sensitive, promoting technological innovation and human capital formation in urban centres, while leveraging renewable energy as a means of inclusive development and income diversification in rural areas.

6. Conclusions and Policy Implications

This study examined the dynamics of renewable energy consumption across 29 European countries from 2000 to 2024, employing Bias-Corrected estimation techniques to analyse the relationship between renewable energy consumption, human development, labour force, and population through global, rural, and urban models. The results indicate that renewable energy consumption is significantly influenced by socio-economic determinants. Higher levels of human development and a larger labour force contribute positively to renewable energy uptake, whereas population growth exerts a negative effect. These findings suggest that the social and demographic composition of a region plays a crucial role in shaping its energy transition trajectory.

These findings underscore the importance of considering socio-economic and demographic contexts when designing energy policies. In particular, they highlight that renewable energy can contribute positively to economic and social development, but the effectiveness of such contributions depends on structural conditions within each region. In addition, the results point to the importance of context-specific strategies, as the blanket use of renewable energy as a tool to reverse rural demographic shifts could lead to unintended outcomes. Nonetheless, the literature suggests that, when carefully implemented, renewable energy initiatives can align with broader energy goals while simultaneously fostering economic and social benefits, particularly through employment generation. This highlights the interconnection between renewable energy policies and the objectives of the Sustainable Development Goals (SDGs), especially SDG 7 (Affordable and Clean Energy), SDG 8 (Decent Work and Economic Growth), and SDG 13 (Climate Action).

From a policy perspective, the findings emphasise that the renewable energy transition should be approached not only as an environmental priority but also as a socio-economic development strategy. Policymakers should integrate renewable energy policies with broader human development and labour market objectives to maximise synergies between clean energy, job creation, and social inclusion. For instance, promoting green skills and vocational training (SDG 4: Quality Education) can strengthen the labour force’s capacity to adapt to the energy transition, while supporting innovation in renewable technologies can foster competitiveness and economic resilience.

Furthermore, European energy and regional policies, such as the EU Green Deal, the Fit for 55 package, and the Just Transition Mechanism, offer valuable frameworks for operationalising these goals. Governments should ensure that renewable energy investments are equitably distributed, particularly in rural and peripheral areas, to mitigate social disparities and prevent unequal access to the benefits of the green transition, thereby advancing SDG 10 (Reduced Inequalities). By encouraging public–private partnerships and community-led renewable energy projects, policymakers can enhance social acceptance, democratise energy access, and create inclusive growth pathways.

A key implication of this research is that renewable energy policy must be territorially sensitive and socially inclusive. Investments in renewable infrastructure should be accompanied by initiatives that improve local infrastructure, promote entrepreneurship, and strengthen institutional capacity to manage the transition effectively. In this sense, renewable energy can act as both a driver and a consequence of sustainable development, reinforcing the virtuous cycle between clean energy adoption, human development, and economic opportunity.

The present work has limitations, and due to data limitations, the analyses are based on some aggregate parameters and estimated parameters, which may mask local heterogeneity and measurement issues. In addition, some relevant factors, such as specific policy measures, energy prices, and technological changes, could not be fully incorporated, and endogeneity concerns may persist despite the dynamic panel approach.

For future research, further examination of the interplay between renewable energy consumption, labour market dynamics, and economic activity is encouraged. Expanding this analysis could clarify the mechanisms through which renewable energy contributes to employment, productivity, and regional resilience, providing additional evidence for policies that balance environmental sustainability with inclusive economic growth.