**1. Introduction**

Nowadays, over half of the world's inhabitants live in cities. The United Nations' prognoses point out that the total population in the world in 2050 will reach 9.31 billion, while the urban population will increase to 6.25 billion, and the urbanization index will be 67.2% [1]. It is largely the civilization advance, together with all the accompanying effects, which has made the population in cities grow dramatically. However, all these aspects and assumptions have consequences as far as the natural environment, the population growth, and the population distribution in particular areas of the globe, especially in cities, are concerned. In the European Union countries, urbanization is progressing continuously, extending into new regions. In the years 2000–2018, its increase was visible in most countries. The urbanization index dropped only in four countries: Slovak Republic, Austria, Cyprus, and Poland. Minor changes (less than 1%) in this respect occurred only in the Baltic countries (Estonia, Latvia, and Lithuania), the Czech Republic, and Belgium (Table A1 in Appendix A). Based on World Bank Statista data, in 2019, 75% of the population lived in cities and the suburbs of the European Union countries, while only 25% lived in rural areas. It is noteworthy that at the same time, a decrease in carbon dioxide emissions can be observed along with the process of growing urbanization.

In most countries, a decreasing carbon dioxide emissions tendency is seen in comparing the emissions in the years 2000 and 2018. For example, Luxembourg, the leader

**Citation:** Jó ´zwik, B.; Gavryshkiv, A.-V.; Galewska, K. Do Urbanization and Energy Consumption Change the Role in Environmental Degradation in the European Union Countries? *Energies* **2022**, *15*, 6412. https://doi.org/10.3390/en15176412

Academic Editors: Junpeng Zhu and Xinlong Xu

Received: 26 July 2022 Accepted: 30 August 2022 Published: 2 September 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

in this ranking in 2000, witnessed significant changes in the field of applied policy concerning emissions reduction. In this region, the estimated decrease in emissions was from 19.7 metric tons per capita in 2000 to 15.3 in 2018 (Table A1 in Appendix A). The visible changes (more than 1%), in this respect, proceed in another direction in the Baltic countries (Lithuania, Latvia, and Estonia). There, an increase in carbon dioxide emissions took place by 1.03, 1.13, and 1.50 metric tons per capita, respectively. However, it must be noted that in most countries, the ongoing changes are an element of the idea of climate neutrality, which is the aim of the European Union for the next decades.

The European Union treats the problem of climate change in a very emphatic way, and it undertakes activities in this direction. Prevention of those changes is one of its priority goals, reflected in the tasks designed for the decades to come, for example, through a reduction in greenhouse gas emissions [2]. The European Green Deal is a strategy for growth, transforming the economic and political union of 27 European democratic countries into places that are neutral to climate. The activities accompanying the major goal refer to significant aspects. Firstly, it is the establishment of a modern, resource-efficient, and competitive economy where there will be no net emissions of greenhouse gases in 2050. Secondly, it is a separation of economic growth from the use of resources. The third aspect refers to guaranteeing the protection and strengthening of neutral capital. Finally, the fourth, but nonetheless very important point, is to ensure citizens' health protection, security, and well-being, which is aimed at protecting them from the environmental effects of climate change.

Considering the formulated aims of climate neutrality, as well as the economic development and the progressing process of urbanization in the European Union countries, our main research goal is to answer the question, is there a long-run relationship between urbanization, energy consumption, economic growth, and the carbon dioxide emissions, and what roles do urbanization and energy consumption play in the concept of the environmental Kuznets curve? This study aims to contribute to this growing area of research by exploring the European Union countries in the period, which covers the accession of new member states from Central Europe. This enlargement needs intensifying cooperation between EU member states, especially in environmental policy. The relationship is tested using the concept of the environmental Kuznets curve, where, apart from carbon dioxide emissions and economic growth, urbanization and final energy consumption are considered. To this aim, the Pedroni and Westerlund panel cointegration tests are used. To estimate the long-run coefficients of the cointegration association, we employed the panel Fully Modified Ordinary Least Squares (FMOLS) test. To test the robustness of the estimation results, we used the Pesaran and Smith Mean Group (MG) estimator, the Pesaran Common Correlated Effects Mean Group (CCEMG), and Augmented Mean Group (AMG) estimators.

The remaining sections of this research are planned as follows. Section 2 presents a brief literature review on the relationship between urbanization, environmental degradation, and economic growth, analyzed within the environmental Kuznets curve (EKC) concept. Section 3 contains the data, model and empirical methodology. The research results and discussion are presented in Section 4. Section 5 concludes the research and provides policy recommendations.

#### **2. Literature Review**

Research on the effect of urbanization on the quality of the environment is frequently conducted using the concept of the environmental Kuznets curve, which appeared at the beginning of the 1990s in the work by Grossman and Krueger [3]. The Authors proved that the scale of environmental pollution is connected with the level of economic development of a given country. In the initial stages of economic development, an increase in the level of pollution related to the exploitation of natural resources also takes place, intending to create welfare. This tendency is reversed after a certain level of income (turning point) is trespassed. Then, the situation changes, and expenditures on environmental protection start to increase. The conclusions drawn by Grossman and Kruger became the basis for creating a model according to which the relationships between economic growth and the emissions of pollutants have an inverted U-shaped curve. In recent years, the popularity of the environmental Kuznets curve, which additionally used different variables, grew. A complex review of the literature in this area can be found, for instance, in Shahbaz and Sinha [4,5], Purcel [6], Koondhar et al. [7], and Xia et al. [8]. This article focuses on research that primarily considers the variables characterizing the urbanization process, urban population, and energy consumption.

It needs to be emphasized that this research was carried out in various regions and states with different levels of economic development, for instance, in emerging economies, developing countries, or developed countries. Most of those studies confirm the relationships defined by the environmental Kuznets curve, but the results of the effect of urbanization on the quality of the environment are not conclusive. For example, Maneejuk et al. [9] analyzed the relationship between GDP per capita, urbanization, financial development, the industrial sector, and the emissions of CO2 for the Association of Southeast Asian Nations (ASEAN), the European Free Trade Association (EFTA), the European Union (EU), Group of Seven (G7), Gulf Cooperation Council (GCC), Mercosur, the North American Free Trade Agreement (NAFTA) and the Organization for Economic Co-operation and Development (OECD) in the years 2001–2016. The findings indicate that the EKC hypothesis is valid in only three out of eight international economic communities, namely, the EU, OECD, and G7. It follows from the research that urbanization, as well as financial development and the industrial sector, increase CO2 emissions, while the use of renewable energy reduces degradation of the environment. In the case of urbanization, statistical significance and the highest positive effect were displayed by ASEAN (0.823), and then by GCC (0.563), Mercosur (0.553), UEU (0.123), and G7 (0.019). In the other groups, the effect of urbanization on CO2 emissions was statistically insignificant.

Similar results were obtained by Wang et al. [1]. The Authors analyzed the effect of urbanization on economic growth and the quality of the environment in the period 1996–2015 based on data from 134 countries. Studies confirmed the occurrence of an inverted U-shaped relationship between economic growth and CO2 emissions for the countries in the lower middle-income group, and a U-shaped relationship for the highincome group of countries. The Authors showed that the emissions of CO2 increased together with increased urbanization. The same direction of the effect of urbanization on carbon dioxide emissions was defined by Sun Y. et al. [10], who conducted research on the Middle East and North African (MENA) economies.

However, it deserves to be pointed out that an increase in urbanization can increase the emissions of carbon dioxide only to a certain level, after which its further progress will reduce these emissions. Such relationships were confirmed in the studies by Gierałtowska et al. [11], who indicated that urbanization has an inverted U-shaped relationship with CO2 emissions in the group covering 163 countries over the period from 2000 to 2016. This relationship can be confirmed by the results of studies obtained by Li and Haneklaus [12], where the Authors showed that increased urbanization decreases CO2 emissions in the group of G7 countries in the years 1979 to 2019, and by Balsalobre-Lorente et al. [13] in the BRICS states in the years 1914–2014. Likewise, studies by Saidi and Mbarek [14], on the effect of urbanization, income, trade openness, and financial development on the carbon dioxide emissions in nineteen emerging economies during 1990–2013, indicate that urbanization decreases CO2 emissions. According to the Authors, this is a powerful argument for politicians and city planners in shaping contemporary policies in those regions.

Previous studies conducted in European countries indicated the opposite results. Based on research carried out in 33 European countries and covering the period 1996–2017, Ali et al. [15] showed that urbanization together with economic growth, export, import, and energy consumption are the main factors that increase environmental degradation. The coefficients associated with urbanization are positive and statistically significant: 0.188 for model I, and 0.011 for model II. At the same time, the Authors point to energy innovation, which should help to reduce the rate of environmental degradation. A similar group of European countries (36) was studied by Wang et al. [16], who indicated a positive and significant effect of urbanization, as well as economic growth and foreign direct investment, on CO2 emissions in the years 2000–2018. A slightly bigger group was examined by Khezri et al. [17]. The results of studies for 43 European countries between 1996 and 2018 also confirmed the relationship defined as the environmental Kuznets curve and urbanization's positive effect on carbon dioxide emissions (coefficients 0.659–0.760).

Comparable results, but for smaller groups of European countries, were obtained by Balsalobre-Lorente et al. [18]. The Authors studied the relationships between GDP per capita, urbanization, foreign direct investment, renewable energy consumption, and CO2 emissions in Portugal, Ireland, Italy, Greece, and Spain, in the years 1990–2019. The study confirmed the relationship between economic growth and CO2 emissions in the inverted U-shaped and N-shaped curves. The urbanization process increases the emissions of CO2 in such a way that an increase in urbanization by 1% increases CO2 emissions within the range from 0.44% to 6.36%, depending on the adopted model. Verbiˇc et al. [19] conducted studies for the countries of South-Eastern Europe in the years 1997–2014. The evidence points to an inverted U-shaped relationship between GDP per capita and the emissions of carbon dioxide in the long run in the whole sample. Short-term estimates evidence the existence of EKC in the inverted U-shape only for Greece and Moldavia. The Authors pointed to a statistically significant positive influence of urbanization on CO2 emissions (coefficient 1.057, FMOLS).

However, not all studies confirm the negative or positive effects of urbanization on the emissions of carbon dioxide. To give an example, no relation between urbanization and carbon dioxide emissions was indicated by Destek et al. [20]. Their research sample comprised Central European countries such as Albania, Bulgaria, Croatia, the Czech Republic, Macedonia, Hungary, Poland, Romania, Slovakia, and Slovenia. The main goal was to find the relationship between CO2 emissions, urbanization, GDP per capita, energy consumption, and trade openness in the years 1991–2011. Studies confirmed the hypothesis of the environmental Kuznets curve in the sample. Results indicate a short-run two-directional causal relation between CO2 and GDP per capita as well as between GDP per capita and energy consumption. There is no relation, however, between urbanization and carbon dioxide emissions. Similar results were obtained by Amin, et al. [21], who point out that urbanization in European countries does have a positive effect on environmental pollution, but it is statistically insignificant. Interestingly, the Authors saw a need to analyze the transport sector as a consequence of the process of urbanization. The Authors argue that transport significantly affects the air quality. They also point out that using renewable energy reduces carbon dioxide emissions from transportation. At the same time, they emphasize that necessary measures should be taken to increase ecological consciousness, especially among the urban population. In this process, it is important to promote environmentally friendly and energy-efficient means of transportation.

Although, the impact of urbanization on the environment in the European Union is related to the fact that some countries have undergone deindustrialization and offshored the environmental effects of their consumption to other countries. Research on industrialization's impact on carbon dioxide emissions mainly focuses on structural changes, where structural changes towards services, usually at higher levels of economic development, improve environmental quality [22–24]. Previous works, including Cherniwchan [25], and Raheem and Ogebe [26], have shown that industrialization is an important determinant in environmental quality changes. Another problem is offshoring the negative ecological impacts, which is often the result of differences in carbon prices in different regions. This phenomenon can lead to the production of energy-intensive goods into "carbon havens", thus creating a "carbon leakage". The observed industry relocation is a significant problem for the European Union and national policymakers [27].

#### **3. Materials and Methods**

We use the model that characterizes the relationships between economic development and the degree of environmental pollution. The first studies on these relationships included those by Grossman and Krueger [3,28], Shafik and Bandyopadhyay [29], Panayotou [30], and Selden and Song [31]. A fast increase in the number of studies led to the formulation of the concept of the environmental Kuznets curve, for example, see Gruszecki and Józwik [ ´ 32]. It assumes a relationship between economic escalation (GDP per capita) and the level of nature contamination (e.g., due to carbon dioxide emissions), mostly in the inverted U-shaped curve. It happens because industrialization is followed by certain negative consequences (for example, pollution of man's natural environment), which grow to a certain point, after which they decrease, even though economic development proceeds. This, on the other hand, follows on from the fact that at a certain stage of advanced economic development, a change can be noticed in the mechanism of demands exhibited by consumers who then, to function, need more services and a cleaner environment. Technological progress also takes place, which does away with the negative effects of contamination of the surrounding world following economic development.

Considering the realization of our research goal, an important problem proves to be the aforementioned relationship described by the environmental Kuznets curve and the observed increase in urbanization and technological progress in the European Union countries. This relationship induces a search for the answer to the following question: is there a long-run relationship between urbanization, energy consumption, economic growth, and carbon dioxide emissions, and what roles do urbanization and energy consumption play in the concept of the environmental Kuznets curve? We use the econometric model with the urbanization variable to answer the questions. The model with the urbanization variable was employed, for example, by Kasman and Duman [33], Ozatac, Gokmenoglu, and Taspinar [34] as well as by Kirikkaleli and Kalmaz [35], Musa et al. [36], and Anwar et al. [37]. Our model will also consider final energy consumption.

The relationship between carbon dioxide emissions, GDP per capita, urban population (urbanization), and final energy consumption per capita is expressed in model I (Equation (1)). We also use model II (Equation (2)) for a robustness check where environmental degradation is proxied as greenhouse gas emissions, expressed in units of CO2 equivalents. All variables are transformed into a natural logarithm format, to avoid multicollinearity issues, reduce the possible outliers from the dataset, as well as overcome the chances of data sharpness and normality [13].

$$\text{LnCO}\_{2\text{it}} = \beta\_0 + \beta\_1 \ln \text{GDP} + \beta\_2 (\ln \text{GDP})^2 + \beta\_3 \ln \text{URB} + \beta\_4 \ln \text{ENC} + \mu\_{\text{it}} \tag{1}$$

$$\text{LnGHG}\_{\text{it}} = \beta\_0 + \beta\_1 \ln \text{GDP} + \beta\_2 (\ln \text{GDP})^2 + \beta\_3 \ln \text{URB} + \beta\_4 \ln \text{ENC} + \mu\_{\text{it}} \tag{2}$$

where β—regression coefficients, CO2—carbon dioxide emissions in metric tons per capita, GHG—greenhouse gas emissions per capita, GDP—gross domestic product per capita (constant 2015 USD), URB—urban population (% of total population), ENC—final energy consumption in tonnes of oil equivalent per capita, μit—error correction term. It should be pointed out that in many scientific studies, the standard measure of urbanization is the share of the population living in urban areas [38].

Before estimating the models, some preliminary tests need to be applied to the panel data. Figure 1 shows the entire research procedure. Initially, we test for cross-section dependence using the Pesaran CD-test [39]. Afterward, in order to discover whether the data of selected variables have stationarity or non-stationarity, we apply the Im–Pesaran– Shin panel unit root test [40] and the second-generation unit root test in the presence of cross-section dependence proposed by Pesaran [41]. Next, we test the long-run relationship (cointegration) among selected variables. To do this, we performed the Pedroni [42,43] and Westerlund [44] panel cointegration tests. These tests are recommended when inter-country convergence is confirmed [45]. The final step of the empirical analysis is estimating the long-run coefficients (elasticities) of the cointegration association concerning urbanization. For that purpose, we employed the panel Fully Modified Ordinary Least Squares test, the Pesaran and Smith [46] Mean Group estimator, the Pesaran [47] Common Correlated Effects Mean Group, and Augmented Mean Group estimators.


**Figure 1.** The model estimation method. Notes: FMOLS—Fully Modified Ordinary Least Squares test; MG—Pesaran and Smith Mean Group estimator; CCEMG—Pesaran Common Correlated Effects Mean Group estimator; AMG—Augmented Mean Group estimator.

Our study sample consists of 28 countries for which we have complete data for the 2000–2018 period (532 observations for each variable). All data were retrieved from the World Bank and Eurostat databases. Table 1 describes the variables and sources of data. Table 2 shows the summary statistics. It should be noted that the differences between the values of the variables are appreciable in our research sample. The CO2 emissions range between 2.97 metric tons per capita in Latvia and 25.67 in Luxemburg, while GDP per capita ranges between USD 3668.65 in Bulgaria and 105,454.7 in Luxemburg. At the same time, we observe considerable differentiation in the urban population, where the smallest values occur in Slovenia (59.7), and the biggest in Belgium (98.0).

**Table 1.** Variables descriptions and sources of data.


Notes: WDI—World Development Indicators; EEA—European Environment Agency.



Notes: CO2—carbon dioxide emissions (metric tons per capita); GDP—GDP per capita (constant 2015 USD); URB—Urban population (% of total population); ENC—Energy final consumption (tonnes of oil equivalent) per capita.

Figure 2 presents changes in aggregated variables for the European Union countries in the years 2000–2018. In the examined period, a decrease in per capita CO2 and greenhouse gas emissions, final energy consumption, and an increase in GDP per capita and urban population occurred. These trends should be assessed in a positive way. Another issue is a growing proportion of the population living in cities (urbanization), which is undoubtedly connected with the demographic changes, economic development, and technological advance observed in economically developed countries.

**Figure 2.** CO2 emissions per capita, greenhouse gas emissions per capita, GDP per capita, urban population, and final energy consumption per capita in the European Union between 2000 and 2018.
