The economic value of sustainability. Real estate market and energy performance of homes

1 Communication of the Commission of the European Parliament and Council to the European Economic and Social Committee and to the Committee of Regions COM (2019) 640 final. 2 Directive 2010/31/EU, 2010 of the European Parliament and of the Council on the energy performance of buildings


INTRODUCTION
The energy transformation of the housing stock is one of the greatest challenges facing the international community.Building stock accounts for 40% of energy consumption and 36% of greenhouse gas emissions in Europe (European Commission, 2020).
The energy transformation of the built heritage involves important actions in terms of energy savings and energy production from renewable sources.The challenge is to meet the carbon neutrality targets introduced by the European Green Deal. 1 The housing market is therefore faced with an unavoidable challenge in the broader process of decarbonising cities. 2  The relationship between energy efficiency and property values is problematic.On the one hand, it can The energy transition and decarbonisation require a major transformation of the housing stock.Responsible for around 40% of fossil fuel consumption and 30% of CO2 emissions, the built heritage must be upgraded in terms of energy performance.The issue is particularly sensitive for Italy, where seven out of ten homes are in the worst energy classes according to the classification introduced by the EU.The research aimed at investigating whether, and to what extent, energy-efficient homes command a premium price on the market compared to less efficient homes.The study focused on three medium-sized cities in northern Italy and developed three hedonic models based on more than 900 ask prices.
The results confirm the stratification of the market in terms of energy efficiency levels.Estimates converge in absolute and relative terms between different cities.Particularly significant is the average gap found between properties in the highest efficiency class (Class A) and those in the lowest efficiency class (Class G), with values of around thirty percentage points, while the gap is around fourteen percentage points if we consider properties in Class D compared to Class G.The conclusions are in line with European studies on the subject, but show a trend towards an increasing gap between the market values of properties with different energy performance classes.

BACKGROUND. FUNDAMENTALS AND MEASURES OF THE PREMIUM PRICE ASSOCIATED WITH THE ENERGY EFFICIENCY OF HOMES
Does the property market differentiate between energy-efficient and inefficient properties?And if so, how?The research question at the heart of this paper has two theoretical references.
The first concerns the possibility of capitalising on savings related to lost operating costs in building management.Buyers of energy-efficient buildings can count on future savings.The cost difference related to lower operating costs can be capitalised and results in a premium price that considers the future savings of the property.Aydin et al. (2020) and Eichholtz et al. (2013) argue that demand accurately values the economic investment and future savings associated with superior energy performance.
There are some exceptions to this argument.Such savings are valued differently by the demand side.In markets where demand pressure is higher, the value difference between properties with different energy performance is lower.Addae-Dapaah and Wilkinson (2020) point to smaller value differentials in the tertiary market in areas of highest demand concentration compared to peripheral areas, confirming Eichholtz et al. (2010) measurements for similar markets.Furthermore, demand may anticipate a gap between actual and declared performance.In the Irish residential market, Coyne and Danny (2021) found that consumption was significantly higher than what was declared in EPC certificates.The gap between actual and declared performance is therefore an important phenomenon (Rajithan et al., 2021), which could influence the extent to which demand values the energy performance of buildings.Finally, demand could be characterised by the behaviour and use patterns of the home such that the performance of the unit is of little relevance (e.g.second homes used only in summer) and thus prove insensitive to energy efficiency.However, these exceptions do not seem to have a significant impact on the overall preferences of the market, which should normally use energy efficiency as a differentiating factor, which in turn should be able to determine a clearly recognisable premium price (Porter, 1985).
The second theoretical reference, on the other hand, concerns expectations regarding the regulation of real estate.As in other markets (Konar & Cohen, 2001), as in the case of the car market (Haščič et al., 2009), the regulator has increasingly imposed constraints in terms of energy and environmental performance.Such adjustments could make it difficult or impossible to sell or rent a property, resulting in a loss of value.
be assumed that the market would capitalise on the superior efficiency of energy-efficient buildings.A change in value is expected because energy-efficient buildings have lower running costs and superior ability to adapt to expected norms.On the other hand, it is possible to assume that energy performance is undervalued because demand may not consider it important.
Deepening the two hypotheses is also relevant to the investment decisions of private operators and affects public policies.In the first case, investment in energy efficiency allows a return linked to the premium price that better-performing buildings can command on the market.In the second case, improving the energy performance of real estate represents a pure cost for the owner.
The Italian scenario has particular characteristics.More than 70% of the country's residential buildings belong to the most energy-intensive classes, Classes E, F, and G of the APE scale (the Italian equivalent of the EPC 3 ).The challenge is considerable in terms of the resources at stake, since the transition to Class D aloneindicated by the European Commission as necessary by 2033 4 -affects a stock of almost 25 million building units. 5 The study aims at analysing property values in the Italian residential market to verify whether and to what extent the energy performance of properties is valued by the market and can therefore be a determining factor in pricing.
The survey is conducted within a precise economic and geographical perimeter.The markets considered are those of three medium-sized cities in the north of the country: Padua, Mestre and Bergamo.The research therefore sought to focus its attention on mediumsized cities, which represent a central element of the Italian territorial structure (IFEL, 2013).
The paper is divided into four sections.The first deals with the main theoretical and empirical references on the subject.The second section presents the methodology used and the data collected.In line with a large body of similar research, the study uses the hedonic price method concerning a dataset covering the three cities.The value analysis models are presented in the third section, while the fourth section deals with their discussion and interpretation.
The economic value of sustainability.

Real estate market and energy performance of homes
The market should therefore highlight a premium price for properties that are protected in the medium and long term from forms of rental and sale restrictions, compared to properties that could find themselves in this situation due to their technological characteristics.In March 2023, the proposal for a directive on the energy performance of buildings 6 , was approved, establishing zero emissions for all new buildings from 2028, while existing buildings will have to reach energy Class E by 2030 and Class D by 2033.The two references now discussed contribute to the theoretical basis for market differentiation between properties characterised by different energy levels.International and national empirical studies confirm this hypothesis.If we focus our attention on the residential market, the reviews by Fuerst et al. (2016), Copiello et al. (2021), Copiello and Donati (2021) and Zhang et al. (2018) trace the main results of research on the premium price associated with energy performance.Although there is no shortage of exceptions (Wahlström, 2016), almost all studies show positive marginal prices associated with residential energy savings.Scientific literature in Europe follows the European Union's Directive on the Energy Performance of Buildings (Directive 2002/91/EC), which makes it mandatory to measure the energy consumption of new and existing buildings, thus providing a common standard for energy classification across the continent.Awareness of these issues is particularly strong in northern European countries.Brounen and Kok (2011) conducted one of the first studies investigating the value differentiation between differently efficient properties in the Netherlands.The research showed the existence of a premium price of 5% related to energy level when considering a Class D home compared to a Class G home.Jensen et al. (2016) studied the effect of EPC classification in the residential segment of the Danish market, highlighting a positive effect of energy efficiency on property values.Other studies limit the scope of the investigation to specific cities.Fuerst et al. (2016) find a 3.5% premium price for the most energyefficient buildings (Classes A, B and C) compared to Class D properties in the city of Helsinki.Fuerst et al. (2015;2016) and the Bio Intelligence Service report (2013) attest to a similar trend for the English market.In the Welsh residential market, Fuerst et al. (2016) found a 12.8% discrepancy in the purchase and sale values of Class A properties compared to Class D properties.Comparable percentages are found in the Irish market, where Hyland et al. (2013) find a gap of 9.3% between Class A and D properties and 10.6% between Class A and Class F/G.In recent years, research has also looked at markets in the Mediterranean region.In Spain, De Ayala et al. (2016) measured the impact of energy efficiency on the price of homes and found a 5.4% premium for properties in Classes A, B, C or D compared to those in Classes E, F or G.A study limited to the city of Barcelona (Marmolejo-Duarte & Chen, 2019) highlights a 7.8% premium price for the most energy-efficient homes (Class A) compared to those classified as nonefficient (Class G).In Italy, research highlights an increase in demand for the energy performance of buildings.The study carried out in Turin by Fregonara et al. (2015) finds a weak relationship between the ask price of a property and its energy efficiency level, a possible symptom of a still low consideration of energy characteristics.The market in the city of Bolzano has been studied twice.In the first study, Morano et al. (2018) compared Bolzano with the city of Bari.In both cities, significant premium prices are recorded, with a higher valuation of energy characteristics in the case of Bolzano.In the second study, Bisello et al. (2020) focus exclusively on the South Tyrolean city and find a 6% premium for Class A properties compared to Class G properties.Properties with superior energy and environmental performance are not only more in line with the decarbonisation and sustainability objectives at the heart of European policy, but also offer economic advantages that the market recognises and values.However, research to date does not seem to agree on the extent and magnitude of the premium price to be attributed to energy performance.The issue is particularly relevant for the residential segment, which represents the largest part of the housing market in Italy and an important part of Italian household wealth (Guerrieri, 2022).

DATA. THE REFERENCE MARKET, CITIES AND DATABASE UNDER INVESTIGATION
The research focuses on the most representative part of the real estate market, the housing market (Guerrieri, 2022).The choice was to exclude large metropolitan areas, but to concentrate on mediumsized cities or specific parts of metropolitan areas.follow the trends of the large metropolitan poles (IFEL, 2013;Antoniucci & Marella, 2016;Micelli & Righetto, 2023).The ask prices obtained concern three northern Italian cities that are representative of medium-sized cities: Padua, Bergamo and Mestre.The latter, although administratively defined as part of the municipality of Venice, has similar economic and demographic characteristics to the other two cities studied.The three cities are characterised by numbers that are representative of the average size of Italian cities (IFEL, 2013).Indeed, the population varies from a minimum of 87,036 inhabitants in the Venetian mainland to a maximum of 209,938 in Padua, with the city of Bergamo recording 121,546 inhabitants. 8The three cities also have a similar administrative rank: Padua and Bergamo are provincial capitals, while Mestre represents the mainland part of Veneto's regional capital.The data collection was made possible thanks to the ask prices of the main buying and selling platforms of the housing market. 9The ask prices collected -901 in all -concern 354 residential units for Padua, 254 for Mestre and 293 for Bergamo.The data were collected in January 2023.
The methodological choice involves the use of the hedonic price model to determine the contribution to the value of the positional and technological characteristics of the properties (Lancaster, 1966;Rosen, 1974;Forte & de' Rossi, 1979;Simonotti, 1993).
The regression used to estimate hedonic prices can assume several functional forms (Kain & Quigley, 1970;Yayar & Demiṙ, 2014).There is no specific guidance in the literature on which functional form is best suited to represent value articulation (Malpezzi, 2002).In this study, the functional form is of the semilogarithmic type, which is particularly popular due to its characteristics, including its ability to minimise the problem of heteroscedasticity (Boza, 2015;Ottensmann et al., 2008).
During the data collection phase, eight variables were collected for each property.In terms of locational characteristics, the macro-area (centre, semi-central and suburbs) was observed, in line with the classification promoted by the Real Estate Market Observatory (OMI) of the Italian Revenue Agency.Additionally, proximity to the relevant local transport infrastructure was observed (Tab.1).
In terms of the typological characteristics of the sample, a distinction was made between individual (villa) and collective (apartment) homes.The form also recorded the size (in square metres of gross floor area) and the number of bathrooms (number of bathrooms in the residential unit).
journal valori e valutazioni  The economic value of sustainability.Real estate market and energy performance of homes  analyses are given in the appendix (Tables A, B and C).The sample analysis shows the preponderance of homes belonging to the lowest energy efficiency classes.In particular, the percentages of the least energy-efficient homes (E, F and G) reach percentages of 79.1%, 79.2% and 66.5% for Padua, Mestre and Bergamo, respectively (Fig. 1).The characteristics of the sample are consistent with the national building stock data, which show a building stock percentage for these three classes, amounting to 70.3%.The descriptive statistics show a homogeneous picture in terms of typological characteristics and state of maintenance.In fact, the majority are multi-family buildings in a good state of maintenance, despite the low energy efficiency highlighted (Tables 2, 3 and 4).The average unit value is € 2,034.85/sqmfor the city of Padua, 1,905.43/sqmfor Bergamo and 1,767.02€/sqm for Mestre (Fig. 2).

METHODS AND ANALYSIS. THE HEDONIC MODELS FOR THE THREE CITIES
The regression model used to estimate hedonic prices uses the natural logarithm of the unit's market value as the unknown variable.The known variables are the locational, typological and technological characteristics of the housing unit.Formally, the model employed is as follows (Fuerst et al., 2016) Eq. ( 1): (1) where -P i is the natural logarithm of the price of a home expressed in €/sqm; -β 0 is the model's constant; -β i is the marginal price of the characteristic; -X i is the numerical value of the variables studied, including the EPC; -e i represents the random error.The regression models for Padua and Mestre show an R 2 of 51.9% and 52.2% and an adjusted and corrected R 2 of 51.3% and 51.2%, respectively.On the other hand, the model for the city of Bergamo shows lower values, with an R 2 of 35.9% and an adjusted R 2 of 34.8%.The reliability of the models was tested using the Fstatistic, whose p-value is less than 0.05 in all three cities (less than 0.001 for Padua, 0.034 for Mestre and 0.024 for Bergamo).The results of the variance analysis (ANOVA) are presented in Table 5.
journal valori e valutazioni No. 34 -2023 8 The economic value of sustainability.

Real estate market and energy performance of homes
The hedonic prices of the characteristics all have a pvalue of less than 0.05, so their reliability can be considered significant.Table 6 highlights the model values and reports the implied marginal prices of the identified characteristics.
It reports the Variance Inflation Factor (VIF), which provides a measure of multicollinearity between the independent variables.The factor allows the detection of any interaction effects present in the model (Bottero et al., 2018).In the three cases considered, a VIF close to one indicates that the variables are weakly correlated and therefore independent.
For two of the three cities studied, the locational characteristics condition the articulation of the values.For Padua and Bergamo, the traditional centre-periphery articulation marks the decline in property values.The case of Mestre is a significant exception, in that the market does not distinguish the traditional hierarchy between the centre and the periphery.Proximity to urban public transport infrastructure is a positional advantage that the models reported in two out of three cases, Padua and Mestre.One city is an exception: in Bergamo, no significant relationship was found between proximity to public transport infrastructure and residential market values.Property types that are adequately equipped record higher values, as indicated by the positive sign for the variable of the number of bathrooms in the cases of Mestre and Bergamo.As maintenance features increase, the unit value of the property also increases in all cities.Finally, in line with the law of diminishing marginal utility, the larger the gross floor area of the property, the lower the unit value.The EPC energy rating is a critical element of value creation in all three urban markets.The model confirms the empirical research cited above on the relevance of superior energy efficiency for value creation.The developed hedonic prices show similar and comparable magnitudes, confirming the robustness of the models.The models development the building energy performance variable in an ordinal form.To reach more robust conclusions regarding the contribution to the value of this intrinsic characteristic of the home, a second model considered energy class membership in the form of multiple dummy variables.The development of models based on variables of the latter type makes the estimation methodology more robust on account of the two alternative classifications of the variable sought and its weight in value creation.
The first includes homes belonging to the efficient classes (EPC A and B).The second includes the intermediate classes (EPC C and D), while the last includes the lowest energy-efficient classes (EPC E, F and G).Membership of each class determines a different hedonic price according to a scheme that is no longer constrained, as before, by the ordinal nature of the variable.
Increasing energy class corresponds to increasing statistically significant hedonic prices (p-value always <0.05): the patterns shown in Table 7 confirm the results of the patterns previously described in Table 6.

DISCUSSION. VALUE DEVIATIONS AND COMPARISON WITH THE LITERATURE
The analysis shows that there are significant differences in value according to the energy quality of the properties, confirming the theoretical premises outlined in the first section and the empirical research developed nationally and internationally.The economic advantage associated with lower operating costs is capitalised in the market value.Added to this is the prospect of marketing the property without constraints and limitations, a characteristic associated with properties of superior energy efficiency.
The difference in value between the energy classes deserves further study to compare the results with other researches carried out in Italy and internationally.
Starting from the models for the three cities, the unit value is calculated for a property in Classes A, D and G.
In the second step, the gap between the identified prices is evaluated in relative and absolute terms.efficient Class A (conservatively the first subclass A), the hedonic prices in absolute value are 802 €/sqm for Padua, 404 €/sqmfor Mestre and 414 €/sqm for Bergamo.In percentage terms, the deviations are 34.18%,26.36% and 29.43%, respectively for the three locations (Fig. 4).
In the three cities, the shift from Class D to Class A shows a significant convergence in percentages, between 12.4% and 15.8%.There is a more pronounced convergence in the premium price between the values of Class G and those of Class A, but this convergence is around thirty percentage points.
The results converge even when the energy variable is reported as a dummy variable.The value gaps highlighted in Figure 5  In other segments of the real estate market, a similar trend of value divergence seems to be confirmed.In the tertiary market, JLL (2023) finds value gaps of more than twenty percentage points between green and traditional buildings, exceeding previous international (Eichholtz et al., 2010) and national (Mangialardo et al., 2019) estimates.Thus, even in segments other than residential, the trend appears to be one of increasing market polarisation between different levels of energy performance.

CONCLUSIONS
The energy transition and decarbonisation require a major transformation of the housing stock.The issue journal valori e valutazioni No. 34 -2023 11  is particularly sensitive for Italy, where seven out of ten homes belong to the least energy-efficient classes, according to the classification introduced by the EU with Directive 31/10.
The study aimed at investigating whether, and to what extent, energy-efficient homes command a premium price on the market compared to less efficient homes.
The study focused on three medium-sized cities in Northern Italy and developed three hedonic models on more than 900 ask prices.The results confirm the stratification of the housing market, with converging estimates of value differences between energy performance classes.Particularly significant is the average discrepancy found between Class A and Class G properties, with values of thirty percentage points, while the discrepancy is smaller -around fourteen percentage points -if we consider Class G compared to Class D properties.Thus, the market differentiates properties according to their energy consumption and environmental sustainability, and in line with what has been argued in other researches (Eichholtz et al., 2010), the gaps seem to increase inversely with the rank and size of the markets considered.Future research will undoubtedly be able to enrich and better articulate the evolution of the market according to the goals of decarbonising the built heritage towards higher standards of sustainability.In particular, the relationship between the vibrancy of markets and the values gaps seems to be an important area of research.The ecological transition could contribute to an increase in value differences between the buildings of large metropolitan centres and those of medium and small centres.Such studies could be carried out using different valuation methods, possibly more suited to the spatial characteristics of the context, such as hedonic models based on geostatistics.
In addition, future research could consider the growing importance along the time axis of the appreciation of the demand for the energy performance of buildings, depending on the ecological transition policies and the collective awareness of energy-saving issues.
The exploration of property investment opportunities certainly seems promising.Property refurbishment can generate significant capital gains, potentially justifying intervention in existing housing stock, with returns that must, however, be commensurate with the costs incurred.Finally, it seems important to consider the design of policies for the energy efficiency of buildings, which can undoubtedly take into consideration a common converging interest between ownership and the public body.Lo scritto è organizzato in quattro sezioni.La prima considera i principali riferimenti teorici ed empirici sul tema.La seconda sezione illustra la metodologia impiegata e i dati acquisiti.In coerenza con un vasto corpo di ricerche analoghe, lo studio impiega il metodo dei prezzi edonici con riferimento a un dataset relativo alle tre città in esame.I modelli di analisi del valore sono presentati nella terza sezione, mentre la quarta affronta la loro discussione e interpretazione.

Figure 1 -
Figure 1 -Energy rating (EPC) of the sampled properties in the three cities and in Italy.

Figure 2 -
Figure 2 -Frequency distribution of ask prices of the three cities.
The market value of a home with the most common characteristics in the sample in Padua is 2,347 €/sqm for energy Class G, 2,719 €/sqm for Class D and 3,150 €/sqm for Class A. The values of homes with the most common characteristics in Mestre are 1,533 €/sqm for a home in Class G, 1,723 €/sqm for a home in Class D and 1,937 €/sqm if the home is classified in Class A. The values of homes in Bergamo are 1,408 €/sqm, 1,602 €/sqm and 1,823 €/sqm for EPC energy performance Classes G, D and A, respectively (Fig.3).The absolute difference between a property in Class G and a property in Class D is 372 €/sqm, 190 €/sqm and 194 €/sqm for the cities of Padua, Mestre and Bergamo, respectively.The absolute differences correspond to very similar percentage increases: 15.84% for Padua, 12.41% for Mestre and 13.77% for Bergamo (Fig.4).On the other hand, if we consider the gap between the least energy-efficient Class G and the most energyjournal valori e valutazioni No. 34 -2023 10

Figure 3 -
Figure 3 -Unit values of homes in Padua, Mestre and Bergamo according to energy class.
show the increasing value of the properties compared to the less efficient classes.More specifically, the gap between inefficient classes E, F and G and moderately-efficient properties classes C and D is 19.60% for the cities of Padua and Mestre and 11.74% for Bergamo.The results also converge for the value gap between the least energy-efficient properties and those in the highest classes (A and B), with percentages of 37.16% for Padua, 37.30% for Mestre and 25.99% for Bergamo.The results of the survey are in line with the academic research on the subject.In particular, the estimated premium prices are in line with the results of empirical studies on the Dutch(Brounen & Kok, 2011), Welsh(Fuerst et al., 2016)  and Irish(Hyland et al., 2013)   housing markets.The three studies report an average premium of 14% for properties in energy class A compared to D. Compared to the Italian studies, the research highlights empirical evidence that is consistent with some studies and contradictory with others.Manganelli et al. (2019)  found similar results to those of this study.In the study of the Bari housing market, the analysis of values is carried out for both central and suburban areas.In the former case, the spread between Class A and Class G homes is 29.41%, while in suburban areas it rises to 41.2%.Less consistent, however, are the values obtained in studies carried out in other cities.Dell'Anna et al.(2019) andBisello et al. (2020) studied the markets of Turin and Bolzano.In both cases, the value deviation is estimated in relation to Class G, the minimum energy class.In Turin, the value deviation from Class A is 6.33%, analogous to the 6.3% found for the city of Bolzano.The difference between the deviations found can be attributed to two factors.The first is related to the dynamism of the property markets considered.Turin is a metropolitan city and Bolzano, despite being a medium-sized city, has a very active market with values of absolute importance in the country.As pointed out byEichholtz et al. (2010), value gaps increase where the real estate market is characterised by less active trading and values.In addition, recent EU positions may have had an effect by further widening the gap of the property values.

Figure 5 -
Figure 5 -Change in value between properties with different energy performance in the cities of Padua, Mestre and Bergamo (energy classification measured with dummy variables).

Figure 4 -
Figure 4 -Change in value between properties in energy classes G and D and between properties in energy classes G and A in the cities of Padua, Mestre and Bergamo.

Table 1 -
Surveyed variables and metrics used to define numeric datasets

Table 5 -
Variance analysis of the three models