Evaluation of Existing Green Office Buildings in Euro-Mediterranean Countries

1. Introduction

1.1. Impact of the Building Sector on Environmental Damage in European Union and Mediterranean Countries

In the European Union (EU), buildings are the largest contributor to environmental pressure, accounting for more than 30% of its total environmental footprint. They are responsible for 42% of the EU’s annual energy consumption and 35% of its annual greenhouse gas emissions, underlining their critical role in climate change [1]. Renovating existing office spaces in the EU rather than building new ones is critical to reducing energy consumption and ensuring a healthy, comfortable and productive indoor climate. Approximately 85% of EU buildings were built before 2000 and 75% have poor energy performance; upgrading this existing stock is essential to reducing the 40% of total EU energy consumption attributed to buildings [2]. As of 2021, the construction industry in Israel accounts for approximately 56% of electricity consumption and approximately 33% of total greenhouse gas emissions [3]. One way to reduce the environmental impact of this sector is to make them more environmentally friendly by retrofitting existing buildings using the Leadership in Energy and Environmental Design (LEED) green building rating system [4]. Gluszak et al. recently demonstrated that the LEED system has become widespread in Europe [5].

1.2. The Evolution of LEED

The LEED system has evolved from version (v) 1.0 to v5 through v2.0, v2.2, v3, v4, and v4.1 [6,7]. LEED 2.0 and later versions provide four certification levels: certified, silver, gold, and platinum. According to LEED v3 2009 and later, certified, silver, gold, and platinum certifications require an overall LEED score of 40-49, 50-59, 60-79, and 80 points or higher, respectively. Currently, LEED v4.1 is the latest version that contains the required sample size to obtain reliable statistical inferences when performing simple linear regression.

Table 1 shows that the LEED system is undergoing significant evolution, moving away from the “prescriptive–descriptive” approach toward a “performance-oriented” approach. The aim of this shift is to ensure that buildings actually perform as designed, particularly in terms of carbon reduction, energy efficiency and productivity.

1.3. LEED Categories and Their Characteristics

Table 2 presents the LEED categories and their characteristics. LEED 2009 v3 and LEED v4/v4.1 share a common foundational structure of credit categories, although more stringent requirements, a greater focus on integrative processes, and a new location-based category were introduced in v4/v4.1. The key differences between LEED 2009 v3 and LEED v4/4.1 in terms of categories are as follows: Location and Transportation was added in LEED v4/v4.1 as a dedicated category (previously part of Sustainable Sites) and the Integrative Process category was introduced to encourage early-stage collaboration [14].

1.4. Characteristics of Key LEED Systems

Between 2000 and 2009, four key environmental rating systems were developed for the LEED: New Construction (LEED-NC), Core and Shell (LEED-C-and-S), Commercial Interiors (LEED-CI), and Existing Buildings (LEED-EB). LEED-NC is designed for new buildings and major renovations [15] and LEED-CI is used for tenant improvements and refurbishments that do not involve the building’s shell or structure [15], while LEED-C-and-S is used for the design and construction of the building’s envelope, structure, and major mechanical, electrical, and plumbing systems, but not the interior fit-out [15]. Finally, LEED-EB focuses on ongoing building operations without any major refurbishments [16]. Originally designed as a program for following up LEED-NC-rated buildings, LEED-EB has become a stand-alone system for owners of existing buildings who wish to obtain eco-certification [17,18].

The author of the present study focused on LEED-EB v4.1 as it places increased emphasis on the performance of existing buildings, recognizing the urgent need to retrofit the vast number of older buildings worldwide, which consume far more energy than modern structures [19].

1.5. LEED v4.1 for Existing Buildings

Table 3 and Table 4 list five performance-based indicators and ten prescriptive indicators, respectively. Pushkar showed that for every office project that achieved LEED-EB v4.1 gold certification, the percentage of binary performance indicators achieved was less than 1% [20]. Therefore, these ten binary prescriptive indicators were excluded from statistical analysis. It should be noted that the maximum number of points for the LEED-EB v4.1 system is 100.

1.6. Research Gap

To examine the relationship between independent LEED categories (parts) and the dependent overall LEED score (whole), most studies have used simple linear correlation [e.g., 21-24], and a few studies have used simple linear regression [e.g., 25,26]. However, all these studies were conducted on LEED-certified buildings in the United States, and the four LEED certification levels, namely certified, silver, gold, and platinum, were combined into one group before correlation analysis or linear regression was conducted [21,22,23,24,25]. Most studies using simple correlation or linear regression were conducted on LEED-NC v3/v4-certified projects [21,22,23,24], and only two studies focused on LEED v4.1 for Healthcare (LEEDE-HC)- and on LEED-EB-certified projects [26,27].

Below are three studies that were taken as the closest analogues to the current study.

• Puskar compared LEED-EB v4.1 gold-certified office projects in six countries in the Euro-Mediterranean region and showed that each of these six countries has a unique LEED certification strategy [20]. Therefore, studying LEED certification strategies in different countries is a pressing issue.
• Pushkar investigated the relationship between the built environment and the LT category in a LEED-EB v4-certified office project in the United States and showed that each LEED certification level—namely certified, silver, gold, and platinum—featured a unique certification strategy [28]. Therefore, studying LEED certification strategies at different levels is a pressing issue.
• Pushkar used simple linear regression to examine the causal relationship between individual LEED performance indicators and the overall LEED score and identified a unique certification strategy for LEED-EB v4.1 gold-certified office projects in the United States [27]. Therefore, studying LEED certification strategies in Euro-Mediterranean countries using simple linear regression is a relevant issue.

1.7. Objective, Novelty and Contribution

The objective of this paper is to examine different LEED certification strategies using simple linear regression between independent LEED performance indicators and the dependent overall LEED score for LEED-EB v4.1 gold-certified office projects in selected European and Mediterranean countries.

The novelty of this study is that each of the six countries exhibits a unique LEED certification strategy, as revealed by simple linear regression analysis. Meanwhile, its practical contribution is that this knowledge will help LEED professionals perform better in these countries.

2. Literature Review

In this section, a statistical analysis of the causal relationship between independent individual LEED categories (i.e., part of the whole) and the dependent overall LEED score (i.e., the whole) is presented. Subsection 2.1 summarizes the results of calculations of the coefficients of determination (R²) for the above relationship. In the literature, if the results of a linear relationship between two variables were presented as a Pearson correlation coefficient (r), they were transformed into R² values. In subsection 2.2, the results of the analysis of the following simple linear regression variables are summarized: the intercept (b₀), the slope (b₁), and R². In quantitative assessments of causal relationships, special attention was paid to critical analysis of the study design.

2.1. Analysis of the Relationship Between Independent LEED Category and Dependent Overall LEED Score in Terms of R²

Table 5 shows the relationship between independent LEED categories and dependent overall LEED scores in terms of R². The EA category exhibits the strongest relationship with the overall LEED score (0.27 ≤ R² ≤ 0.63), while the WE and MR categories exhibit the weakest relationship with the overall LEED score (0.04 ≤ R² ≤ 0.15) and the following categories exhibit an intermediate relationship with the overall LEED score (0.08 ≤ R² ≤ 0.21): LT, SS, IEQ, and RP. The analyzed studies [21,22,23,24,25] have at least three common features: (i) the relationship between individual LEED categories and the overall LEED score is statistically significant (p < 0.001), (ii) the analysis of this relationship was conducted on LEED-certified projects in the United States, and (iii) the four LEED certification levels (certified, silver, gold, and platinum), which are four independent groups each with its own unique LEED certification strategy, were combined into one group, which was analyzed using significance tests.

2.2. Analysis of the Relationship Between Independent LEED Category and Dependent Overall LEED Score in Terms of Simple Linear Regression

Before interpreting the LEED results presented in Table 6, it is necessary to clarify LEED terminology. LEED-NC v4.1 features the EA and MR categories as in v3 and 4, while LEED-EB v4.1 features the “energy” and “waste” performance indicators. In this case, the term “energy” is analogous to the term “EA category”, and the term “waste” is analogous to the term “MR category”.

Table 6 shows that the group of LEED gold-certified projects significantly outperformed the combined group of LEED-certified projects, which included four certification levels (certified, silver, gold, and platinum), with regard to linear regression parameters: b₀, b₁, and R². In both cases [26,27], the relationship between the independent individual LEED-HC v4.1 categories or LEED-EB v4.1 performance indicators and the dependent overall LEED score was statistically significant (p ≤ 0.025).

Table 6. A summary of the intercept (b₀), slope (b₁), and coefficient of determination (R²). A simple linear regression between dependent overall LEED scores and independent LEED performance indicators.

LEED	Dependent variable	Independent variable	Regression variable			Ref.
			b0	b1	R2
LEED-HC v4.1-certified projects in the United States	Overall LEED	EA	8.89	0.16	0.36	[26]
LEED-EB v4.1 gold-certified office projects in New York City		Energy	48.13	0.78	0.47	[27]
LEED-EB v4.1 gold-certified office projects in Washington, D.C.			44.96	0.96	0.68
LEED-HC v4.1-certified projects in the United States		MR	6.78	0.04	0.11	[26]
LEED-EB v4.1 gold-certified office projects in New York City		Waste	54.94	1.72	0.19	[27]
LEED-EB v4.1 gold-certified office projects in Washington, D.C.			53.46	1.97	0.42

Table 6. A summary of the intercept (b₀), slope (b₁), and coefficient of determination (R²). A simple linear regression between dependent overall LEED scores and independent LEED performance indicators.

LEED	Dependent variable	Independent variable	Regression variable			Ref.
			b0	b1	R2
LEED-HC v4.1-certified projects in the United States	Overall LEED	EA	8.89	0.16	0.36	[26]
LEED-EB v4.1 gold-certified office projects in New York City		Energy	48.13	0.78	0.47	[27]
LEED-EB v4.1 gold-certified office projects in Washington, D.C.			44.96	0.96	0.68
LEED-HC v4.1-certified projects in the United States		MR	6.78	0.04	0.11	[26]
LEED-EB v4.1 gold-certified office projects in New York City		Waste	54.94	1.72	0.19	[27]
LEED-EB v4.1 gold-certified office projects in Washington, D.C.			53.46	1.97	0.42

3. Materials and Methods

3.1. Study Design and Data Collection

The following conditions were considered when developing the study: LEED-certified projects must belong to the same region (e.g., Euro-Mediterranean), LEED system (e.g., LEED-EB), version (e.g., v4.1), certification level (e.g., gold), and building type (e.g., office). Office building projects that received LEED-EB v4.1 gold certification differed only by country and not by region [20].

To justify the minimum sample size when using simple linear regression to analyze environmental performance indicators in LEED-EB v4.1-certified projects, the author of this study has suggested that LEED data could be considered as part of an environmental field study [29]. Therefore, according to generally accepted recommendations for the use of simple linear regression in ecological field studies, the minimum sample size (n) is n = 15–20 [30]. For this study, the minimum sample size was n = 20.

Table 7 lists six countries in the Euro-Mediterranean region, and as of 7 January 2026, at least 20 LEED-EB v4.1 gold-certified office projects in each country were identified from two comprehensive databases [31,32]. In a previous study, six of the twenty Euro-Mediterranean countries, namely Sweden, Italy, Israel, Spain, Germany and Ireland, were selected based on the number of LEED-certified projects required to draw reliable statistical conclusions. This study is a continuation of this previous study [20].

3.2. Statistical Analysis of Linear Regression

The following variables were used to interpret the results of simple linear regression: p-value, b₀, b₁, and R². If the relationship between the dependent variable and one independent variable was statistically significant (p ≤ 0.05), then the values of b₀, b₁ and R² were used to interpret the linear regression. If the relationship between the dependent variable and one independent variable was statistically insignificant (p > 0.05), then the values of b₀, b₁ and R² were not used to interpret the linear regression. In all cases, a two-tailed p-value was used. MATLAB 2024a was used to process data from LEED-certified projects [33].

4. Results

4.1. Sweden

Figure 1 and Table 8 show that in Sweden, the indicators “water”, “energy”, “waste” and “IEQ” show a positive and statistically significant causal relationship with “overall LEED”, while the “transportation” indicator is not correlated with “overall LEED”. Linear regression showed that each additional point in the “water”, “energy”, “waste”, and “IEQ” indicators contributes approximately 1.07, 1.04, 0.98, and 1.19 points to the “overall LEED” score, respectively. According to the R² value, 29%, 38%, 13%, and 26% of the variance in the “water”, “energy”, “waste”, and “IEQ” variables is predicted by the “overall LEED” variable in the model, respectively.

4.2. Italy

Figure 2 and Table 9 show that, in Italy, the “transportation”, “energy”, and “IEQ” indicators show a positive and statistically significant causal relationship with “overall LEED”, while the indicators “water” and “waste” are not correlated with the “overall LEED”. Linear regression showed that each additional point in the “transportation”, “energy”, and “IEQ” indicators contributes approximately 1.40, 0.97, and 0.89 points to the “overall LEED” score, respectively. According to the R² value, 17%, 46%, and 24% of the variance in the “transportation”, “energy”, and “IEQ” variables is predicted by the overall LEED variable in the model, respectively.

4.3. Israel

Figure 3 and Table 10 show that, in Israel, the “water”, “energy”, and “waste” indicators show a positive and statistically significant causal relationship with “overall LEED”, while the “transportation” and “IEQ” indicators are not correlated with the “overall LEED”. Linear regression showed that each additional point in the “water”, “energy”, and “waste” indicators contributes approximately 1.06, 0.90, and 2.66 points to the “overall LEED” score, respectively. According to the R² value, 28%, 53%, and 56% of the variance in the “water”, “energy”, and “waste” variables is predicted by the overall LEED variable in the model, respectively.

4.4. Spain

Figure 4 and Table 11 show that, in Spain, the “transportation”, “water”, and “energy” indicators show a positive and statistically significant causal relationship with “overall LEED”, while the “waste” and “IEQ” indicators are not correlated with the “overall LEED”. Linear regression showed that each additional point in the “transportation”, “water”, and “energy” indicators contributes approximately 1.22, 0.60, and 0.76 points to the “overall LEED” score, respectively. According to the R² value, 20%, 16%, and 40% of the variance in the “transportation”, “water”, and “energy” variables is predicted by the overall LEED variable in the model, respectively.

4.5. Germany

Figure 5 and Table 12 show that, in Germany, the “water”, “waste”, and “IEQ” indicators show a positive and statistically significant causal relationship with “overall LEED”, while the “transportation” and “energy” indicators are not correlated with the “overall LEED”. Linear regression showed that each additional point in the “water”, “waste”, and “IEQ” indicators contributes approximately 1.29, 2.11, and 0.67 points to the “overall LEED” score, respectively. According to the R² value, 35%, 39%, and 27% of the variance in the “water”, “waste”, and “IEQ” variables is predicted by the overall LEED variable in the model, respectively.

4.6. Ireland

Figure 6 and Table 13 show that, in Ireland, the “waste” and “IEQ” indicators show a positive and statistically significant causal relationship with “overall LEED”, while the “transportation”, “energy”, and “water” indicators are not correlated with the “overall LEED”. Linear regression showed that each additional point in the “waste” and “IEQ” indicators contributes approximately 3.43 and 0.90 points to the “overall LEED” score, respectively. According to the R² value, 42% and 47% of the variance in the “waste” and “IEQ” variables is predicted by the overall LEED variable in the model, respectively.

The key components of variation and the regression coefficients from the statistically significant linear regression results are presented in the ANOVA (Analysis of Variance) regression tables and regression coefficient tables, respectively (Appendix A, Table A1, Table A2, Table A3, Table A4, Table A5, Table A6, Table A7, Table A8, Table A9, Table A10, Table A11, Table A12, Table A13, Table A14, Table A15, Table A16, Table A17, Table A18, Table A19, Table A20, Table A21, Table A22, Table A23, Table A24, Table A25, Table A26, Table A27, Table A28, Table A29, Table A30, Table A31, Table A32, Table A33, Table A34, Table A35 and Table A36). Statistically insignificant linear regression results were not included in these tables.

5. Discussion

5.1. LEED Certification Strategy in Euro-Mediterranean Countries

Table 14 shows that each of six countries has a unique LEED certification strategy. In Sweden, for LEED-EB v4.1 gold-certified office projects, there was a statistically significant causal relationship between four of the five performance indicators and the overall LEED score. In Italy, Israel, Spain, and Germany, for the same office projects, there was a statistically significant causal relationship between three of the five performance indicators and the overall LEED score. Finally, in Ireland, there was a statistically significant causal relationship between two of the five performance indicators and the overall LEED score for LEED-EB v4.1 gold-certified office projects. It can be hypothesized that the wider range of causal relationships between independent LEED performance indicators and the resulting overall LEED score is associated with the greater diversity of LEED certification strategies [27]. Although the number of causal pairs is the same for Italy, Israel, Spain, and Germany (three pairs in each country), they each have unique sets of pairs. Thus, this comparison showed that LEED certification strategies in Euro-Mediterranean countries differ significantly from those in cities in the United States [27].

Table 14 also shows that statistically significant differences are observed in the causal relationship between the transportation indicator and the overall LEED score in two countries, namely Italy and Spain, while this difference is not statistically significant in four countries. The relationship between the “water” indicator and the “overall LEED” score showed statistically significant differences in four countries, namely Sweden, Israel, Spain and Germany, while no statistically significant differences were found in two countries, namely Italy and Ireland. There were statistically significant differences in the relationship between the “energy” indicator and the “overall LEED” score in four countries, namely Sweden, Italy, Israel, and Spain, while these differences were not statistically significant in two countries, namely Germany and Ireland. The relationship between the “waste” indicator and the “overall LEED” score showed statistically significant differences in four countries, namely Sweden, Israel, Germany, and Ireland, while no statistically significant differences were found in two countries, namely Italy and Spain. Finally, the relationship between the “IEQ” indicator and the “overall LEED” score exhibited statistically significant differences in four countries, namely Sweden, Italy, Germany, and Ireland, while no statistically significant differences were found in two countries, namely Israel and Spain.

5.2. Comparison of the b₀, b₁, and R² Values with Literature Data

Table 15 presents simple linear regression variables from recent and current studies on LEED-EB v4.1 gold-certified office projects. A recent study analyzed LEED-EB v4.1 gold-certified office projects in two US cities: New York City and Washington, D.C. This study found that statistically significant relationships were observed in only two out of five independent LEED performance indicators and the dependent overall LEED score, namely “energy” and “waste” [27]. In the current study, two of the six countries, namely Sweden and Israel, exhibited a similar pattern: the “energy” and “waste” indicators exhibited a statistically significant relationship with the overall LEED score. Comparative analysis of regression coefficients showed that only Washington, D.C. and Israel had similar regression coefficients.

6. Conclusions

The objective of this study was to evaluate LEED certification strategies by analyzing the causal relationships between five independent LEED performance indicators and the dependent overall LEED score using simple linear regression. The following conclusions were drawn:

• Each of the six countries analyzed, namely Sweden, Italy, Israel, Spain, Germany and Ireland, demonstrated a unique LEED certification strategy for LEED-EB v4.1 gold-certified office projects.
• This is the first case in the literature where the causal relationship between “energy” and “overall LEED” in LEED-EB v4.1 gold-certified office projects in Germany and Ireland was found to be statistically insignificant (R² = 0.04 and p = 0.359; R² = 0.13 and p = 0.112, respectively).
• In contrast, in Sweden, Italy, Israel and Spain, the causal relationship between “energy” and “overall LEED” was statistically significant (R² = 0.38, 0.46, 0.53 and 0.40 at p < 0.001 in all cases, respectively).

7. Limitations and Future Research Directions

This study describes different LEED certification strategies in Euro-Mediterranean countries without taking into account external factors. Future research should examine the causal relationships among green building policies [34], climate zones [35], urban planning [36] and individual LEED performance indicators in each Euro-Mediterranean country.