1# RECORD c457882224ad7235eecc5a3f2b95232d76a155e72# TITLE: Unleashing the hidden green value: assessing the impact of energy certification and environmental amenities on real estate through hedonic modelling3# YEAR: 2025.0 VENUE: Aestimum DOI: 10.36253/aestim-174184# AUTHORS: Carolina Bonardi Pellizzari; T. Tempesta; C. Franceschinis; Mara Thiene; D. Vecchiato5# PAGES: 16 CANDIDATE TABLE PAGES (1-based): [9]6# ABSTRACT: Rapid population growth and urbanization have intensified energy demand and climate vulnerability in cities, elevating the importance of urban green spaces and energy-efficient properties, which impact real estate values. This study uses a hedonic pricing model to assess how environmental amenities and energy efficiency influence property prices. We analysed sales contracts from 2022–2023 in Padua (Italy), enriched with spatial data on key environmental factors. Findings reveal a notable price premium associated with proximity to green spaces and waterways: for each hectare of park area, properties located within 100 m experience a 5.4% price increase, while each additional 100 meters from a waterway reduces apartment value by 3.7%. Additionally, apartments in energy class A or higher command a 30% higher price per square meter with respect to other energy classes. This research offers valuable insights into how urban green spaces and energy efficiency shape real estate values in rapidly urbanizing settings.789===== PAGE 1 (score 13, numeric-paren 0, table-words 0, var-words 13) =====1011Aestimum 87: 71-86, 202512Firenze University Press 13www.fupress.com/ceset14A E S T I M U M15CENTRO STUDI DI ESTIMO E DI ECONOMIA TERRITORIALE - Ce.S.E.T.16ISSN 1592-6117 (print) | ISSN 1724-2118 (online) | DOI: 10.36253/aestim-17418 17Citation: Bonardi Pellizzari, C., 18Tempesta, T., Franceschinis, C., 19Thiene, M., & Vecchiato, D. (2025). 20Unleashing the hidden green value: 21assessing the impact of energy 22certification 23and 24environmental 25amenities on real estate through 26hedonic modelling. Aestimum 87: 71-86. 27doi: 10.36253/aestim-17418 28Received: March 4, 202529Accepted: June 16, 202530Published: December 31, 202531© 2025 Author(s). This is an open 32access, peer-reviewed article pub-33lished by Firenze University Press 34(https://www.fupress.com) and distrib-35uted, except where otherwise noted, 36under the terms of the CC BY 4.0 37License for content and CC0 1.0 Uni-38versal for metadata.39Data Availability Statement: The 40datasets used and/or analyzed during 41the current study are available from 42the corresponding author on reason-43able request.44Conflicts of Interest: The authors 45declare no conflict of interest. The 46funders had no role in the design of 47the study; in the collection, analyses, 48or interpretation of data; in the writing 49of the manuscript, or in the decision to 50publish the results.51ORCID:52CBP: 0000-0002-6019-968653TT: 0000-0002-6445-474454CF: 0000-0001-5641-782355MT: 0000-0003-3265-542656DV: 0000-0002-4652-396157Unleashing the hidden green value: assessing 58the impact of energy certification and 59environmental amenities on real estate through 60hedonic modelling61Carolina Bonardi Pellizzari*, Tiziano Tempesta, Cristiano France-62schinis, Mara Thiene, Daniel Vecchiato63Department of Land, Environment, Agriculture and Forestry (TESAF), University of 64Padua, Italy65E-mail: carolina.bonardipellizzari@phd.unipd.it, tiziano.tempesta@unipd.it, cristiano.66franceschinis@unipd.it, mara.thiene@unipd.it, daniel.vecchiato@unipd.it67*Corresponding author68Abstract. Rapid population growth and urbanization have intensified energy demand 69and climate vulnerability in cities, elevating the importance of urban green spaces and 70energy-efficient properties, which impact real estate values. This study uses a hedon-71ic pricing model to assess how environmental amenities and energy efficiency influ-72ence property prices. We analysed sales contracts from 2022–2023 in Padua (Italy), 73enriched with spatial data on key environmental factors. Findings reveal a notable 74price premium associated with proximity to green spaces and waterways: for each hec-75tare of park area, properties located within 100 m experience a 5.4% price increase, 76while each additional 100 meters from a waterway reduces apartment value by 3.7%. 77Additionally, apartments in energy class A or higher command a 30% higher price per 78square meter with respect to other energy classes. This research offers valuable insights 79into how urban green spaces and energy efficiency shape real estate values in rapidly 80urbanizing settings.81Keywords: building’s energy performance, hedonic model, urban environment.82JEL codes: Q49, Q51, R30.831. INTRODUCTION84The world’s population is three times larger than it was in the mid-twen-85tieth century, growing from 2.5 billion in 1950 to 8 billion in 2022, and it is 86projected to reach up to 9.7 billion in 2050 (United Nations, 2022). Around 8755% of this total currently lives in urban areas and this share is projected to 88reach 68% by 2050 (United Nations, 2019). 89The increasing number of people in the cities coupled with the conse-90quences of climate change, such as higher temperatures, floods, and exces-91sive drought is increasing the vulnerability of the population living in urban 929394===== PAGE 9 (score 51, numeric-paren 11, table-words 10, var-words 17) =====95967997Assessing the impact of energy certification and environmental amenities on real estate through hedonic modelling98built apartments typically belong to the highest ener-99gy classes. However, the energy class is in some ways a 100proxy for the age of the property and therefore, by omit-101ting this last characteristic, the effect of the energy class 102can be overestimated. From the estimated model in 103which the age of the property is inserted, it emerges that: 104(i) apartments belonging to energy class A or greater 105increase the price per square meter by 30 %. We reduced 106the energy classes in the model to two classes (>=A, all 107other classes) because in further exploratory analyses we 108found that apartments belonging to other energy classes 109did not have a statistically significant effect on the price.110As to the effect of age, it is interesting to note that in 111the city of Padua this variable presents a parabolic trend 112(Figure 2 and Table 3). Including both the linear and 113squared terms allows us to account for the possibility 114that price does not decline uniformly with age. Instead, 115the relationship may follow a U-shaped curve (Figure 2), 116where price initially decreases with age (captured by the 117negative linear coefficient in Table 3), but beyond a cer-118tain point, the effect of age becomes less negative or even 119slightly positive (captured by the positive squared term 120in Table 3). This pattern can reflect architectural or his-121torical value in older buildings, or improvements made 122over time. More specifically, it causes the price of the 123property to decrease until the age of 67, after which the 124price per square meter increases. It can be deduced that, 125at least generally, the oldest properties built before the 126Second World War are more appreciated by the market 127for their architectural characteristics or for their location 128closer to the city centre.129The last intrinsic characteristic analyzed, the year 130of installation of the heating system, presented a posi-131tive effect on the price, regardless of the energy class to 132which it belongs (0.69% for each additional year).133Regarding the extrinsic non-environmental char-134acteristics, the model highlights that proximity to some 135services increases the price of the property. The price 136decreases by 1.23% as the distance from the train sta-137tion increases by 100 m. The number of schools within 138200 m increases the unit price by 12.2% and the number 139of supermarkets by 9.6%. These estimates align with the 140urban income theory, which posits that the willingness to 141pay for a property also depends on future costs associated 142with accessing the diverse range of essential services.143Concerning the environmental characteristics, two 144variables are significantly correlated to the price: the 145area occupied by urban parks within a radius of 100 m 146and from 100 to 200 m and the distance from water-147ways. When it comes to the urban parks variable, it 148should be noted that their effect is statistically significant 149up to a distance of 200 m. It can also be seen how the 150effect decreases as the distance of public greenery from 151the home increases: one hectare of park located less than 152100 m from the home increases the price by 5.4%, this 153premium drops to 3.9% for parks placed at a distance 154between 100 and 200 m. The distance from waterways 155also has a significant effect on the price of houses: for 156Table 3. Hedonic model results. Dependent variable: log of apartment prices per square meter (€/m2).157Variables158Coefficients159p-value160Price (€/m2)161% variation162Constant163-5.651164Apartment surface (m2)165-0.00347166****167-0.34168Energy class A1690.265170**17130.34°172Age of the apartment173-0.0166174****175Squared age of the apartment1760.000125177****178Year in which the heating system was installed1790.00691180***1810.69182Bathrooms number1830.173184***18517.35186Elevator1870.161188****18917.42°190Train station distance (Km)191-0.123192**193-12.32194Number of schools within a radius 200 m1950.122196**19712.18198Number of supermarkets within a radius 200 m1990.0965200***2019.65202Urban parks area (ha) within a radius of 100 m2030. 0541204***2055.41206Urban parks area (ha) between 100 and 200 m2070. 0393208**2093.93210Waterway distance (Km)211-0. 366212****213-36.6214Mediation2150.11860216**21712.59°218Observations = 222 / R-squared = 0.551 / Adjusted R-squared = 0.521 / St. Error of the model = 0.378.219° Calculated using Equation (4).220Note: * p<0.10, ** p<0.05, *** p<0.01, **** p<0.001.221222223===== PAGE 10 (score 10, numeric-paren 0, table-words 0, var-words 10) =====22422580226Carolina Bonardi Pellizzari et al.227every additional 100 m of distance from the waterway, 228the apartment price is reduced by 3.7%.229Finally, the model demonstrates that a subjective 230characteristic unrelated to the property itself or the pro-231vision of economic, social, and environmental services 232can influence house prices. Specifically, when an apart-233ment was sold through a real estate agency (mediation), 234the premium was on average 12.6% higher. This could be 235due to Italian legislation, which mandates that the price 236declared in the sales deed must match the actual price 237paid if a real estate agent is involved. Alternatively, it is 238possible that real estate agents are able to increase the 239sales price or that they predominantly handle higher-240priced properties, resulting in a higher average unit price.2415. DISCUSSION242From the analysis of the transaction contracts 243and geospatial data, some clues regarding the effect of 244environmental characteristics on property prices have 245emerged. The findings indicate a statistically significant 246influence of both energy efficiency and extrinsic envi-247ronmental factors on apartment prices in Padua, repre-248sented by price premiums. 249First, the price premium of 30% found for best-per-250forming apartments (EPC=A) is consistent with the liter-251ature which has been showing the increased importance 252of energy performance on real estate. The estimated 253price premium found in this study is somewhat similar 254to the one of Massimo et al. (2022) in the Italian munici-255pality of Reggio Calabria where a 41.5% premium was 256found for residential properties presenting EPC A or B 257compared to other energy classes. Using an intermediary 258baseline (EPC=D) in the same city explored in the pre-259sent study, Copiello and Coletto (2023) found a premium 260price of 55% for apartments belonging to EPC=A4 and 261a decrease of 29% for those with EPC=G. Price premi-262ums were also found in other parts of Italy, such as in 263the municipality of Bolzano, where a premium of 6% 264was found for properties moving from the worst ener-265gy performance class (G) to the best (A) (Bisello et al., 2662020). In Bari, studies using transaction prices identified 267premiums from 27% to 29.41% in properties presenting 268EPC>A (Manganelli et al., 2019; Morano et al., 2020) 269and a decrease of 27% for the worst-performing dwell-270ings (EPC=G) (Morano et al., 2020).271In the context of examining the influence of energy 272performance on the real estate market, a relevant obser-273vation arising from our study is the correlation identified 274between the apartment’s age and its EPC rating. With-275in our sample, newly built apartments exhibit superior 276energy performance, introducing the potential for con-277trasting conclusions. Divergent approaches exist in the 278literature: some studies incorporate both age and EPC 279in their models (e.g. Barreca et al. (2021); Manganelli et 280al. (2019); Morano et al. (2020)), while others opt for a 281singular variable, with EPC serving as a proxy for addi-282tional factors such as a building’s age or simply because 283other attributes are not significantly affecting price (see 284Bisello et al. (2020); Copiello and Coletto (2023); Mas-285simo et al. (2022)). Certain authors consider this correla-286tion problematic, as omitting variables correlated to the 287EPC may lead to overlooking the impact of the selected 288variable on price (Olaussen et al., 2017). 289Regarding extrinsic environmental variables, our 290findings suggest a positive effect of the proximity to 291water, illustrated by a 3.7% decrease in price for every 292additional 100 meters from the waterway, which aligns 293with the research of Bonetti et al. (2016) who identified 294an even stronger price decrease in Milan (Italy), of 7.4% 295for houses for an additional meter of distance from the 296canal. Taking into consideration the effect of blue spac-297es in general, our findings are somewhat aligned with 298Anderson and Vahe (2018), who identified a premium 299price of 0.9% for those properties located within 200 m 300from a small blue space in the UK. In the Netherlands, 301a 1.91% discount in price was found for properties locat-302ed 25-50 m from the nearest water when compared to 303properties located less than 25 m away and the discount 304increased to -4.04% for properties positioned 200-400 305m from the nearest water body (Koster, 2022). Another 306study in the Netherlands estimated a premium price of 3077% for a house located within 1000 m from a lake, and 308this premium got up to 28% if the house was by the lake 309with a nice view (Luttik, 2000). 310Finally, the effect of green urban areas is evidenced 311in our results by a 5.4% premium for an additional hec-312tare of urban park located within 100 m of the apart-313ment (and 3.9% for a distance between 100 and 200 m) 314aligns with existing literature regarding size and prox-315imity of green features. The positive correlation between 316the size of the closest green parks and property value 317has been consistently supported by various studies. For 318instance, an additional 1% of green space within 500 319m of apartments in Poland increased the sale price by 3203.95 PLN per square meter (Czembrowski and Kronen-321berg, 2016). Similarly, in Finland, an additional 1 km 322in the distance to the nearest forested area resulted in a 323decrease of 5.9% in the price of the dwelling (Tyrväin-324en and Miettinen, 2000). In Costa Rica, Piaggio (2021) 325reported a 4% premium price for each additional 1% 326increase in size of the closest neighborhood park. Simi-327larly to our findings, the authors also observed that dis-328