DEGREE PROJECT
REAL ESTATE AND CONSTRUCTION MANAGEMENT BUILDING AND REAL ESTATE ECONOMICS
MASTER OF SCIENCE, 30 CREDITS, SECOND LEVEL STOCKHOLM, SWEDEN 2020
Environmental
Certifications in the Office Segment
Does the Swedish Market Reward
Certified Properties with a Higher Price?
Amanda Grafström & Hanna Käck
TECHNOLOGY
DEPARTMENT OF REAL ESTATE AND CONSTRACTION MANAGEMENT
ROYAL INSTITUTE OF TECHNOLOGY
DEPARTMENT OF REAL ESTATE AND CONSTRUCTION MANAGEMENT
Master of Science Thesis
Title
Authors Department
Master Thesis number Supervisor
Keywords
Environmental Certifications in the Office Segment - Does the Swedish Market Reward Certified Properties with a Higher Price?
Amanda Grafström & Hanna Käck Real Estate and Construction Management TRITA-ABE-MTB-20496
Åke Gunnelin
Environmental certification, transaction price, price premium, office properties, Swedish real estate market
Abstract
This thesis aims to explore if and why there exists a price premium for environmentally certified office properties on the Swedish market. The purpose is to contribute with knowledge within the research field of certified properties and how they are valued by the market. In order to answer the research question, regressions of office transaction data have been performed, of which the main model examined whether environmentally certified office properties generate a price premium compared to non-certified properties. Two additional models were analysed,
investigating if certain certifications affect the price premium more than others. The transaction data was provided by the real estate consultancy firm Cushman & Wakefield. Furthermore, in contrast to previous research, a cost variable was included in each model. However, in order to capture the significance of the parameter, regressions excluding the cost variable were as well performed. The result of the analysis suggests that environmentally certified office properties are sold at a 14-26% higher price compared to equal properties which are not certified. In addition, certified properties with BREEAM and LEED appear to be sold at a 28-30% respectively 30-31%
higher price than GreenBuilding certified properties. The cost variable did not seem to have any larger impact of the price. Most likely, a large part of the generated price premium is a result of aiming for branding benefits which are received by obtaining environmentally certified
properties.
Acknowledgement
This master thesis has been written during the spring semester of 2020 and is the final
examination of the engineering program Civil Engineering and Urban Management, within the master Real Estate and Construction Management at the Royal Institute of Technology.
We would like to give a big thank you to the real estate consultancy firm Cushman & Wakefield, for the provided information and guidance which without, this thesis would not have been
possible to realise.
At last, we would like to express our gratitude towards our supervisor Åke Gunnelin for support and valuable thoughts.
Stockholm, May 2020
Amanda Grafström & Hanna Käck
Examensarbete
Titel
Författare Institution
Examensarbete Master-nivå Handledare
Nyckelord
Miljöcertifierade fastigheter i kontorssegmentet – Belönar svenska marknaden certifierade
fastigheter med ett högre pris?
Amanda Grafström & Hanna Käck Fastigheter och byggande
RITA-ABE-MTB-20496 Åke Gunnelin
Miljöcertifieringar, transaktionspris, prispremie, kontorsfastigheter, svenska fastighetsmarknaden
Sammanfattning
Målet med examensarbetet är att undersöka om och varför det existerar en prispremie för miljöcertifierade kontorsfastigheter som är lokaliserade på den svenska fastighetsmarknaden.
Syftet är att bidra med kunskap inom forskningsområdet certifierade fastigheter och förstå hur de värderas av den svenska marknaden. För att kunna besvara forskningsfrågorna har
regressionsanalyser genomförts för data innehållande transaktioner av kontorsfastigheter.
Huvudmodellen har undersökt om miljöcertifierade fastigheter genererar en prispremie jämfört med icke-certifierade fastigheter. Utöver huvudmodellen har ytterligare två modeller analyserats som undersöker huruvida specifika certifieringar påverkar prispremien. Använd transaktionsdata har inhämtats från fastighetskonsultsföretaget Cushman & Wakefield. Jämfört med tidigare studier har en kostnadsparameter inkluderats i varje modell. För att kunna förstå variabelns signifikans och påverkan har även analyser som exkluderar parametern genomförts. Resultatet av regressionen antyder att miljöcertifierade kontorsfastigheter säljs till ett pris som är 14-26 % högre än jämförbara kontorsfastigheter utan certifiering. Vidare verkar BREEAM- och LEED- certifierade fastigheter säljas till 28-30 % respektive 30-31 % högre pris än GreenBuilding- fastigheter. Kostnadsparametern verkar ej ha en signifikant påverkan på prispremien. Med största sannolikhet utgör strävan efter brandingfördelar en stor del av prispremien vilka erhålls om fastighetsägaren innehar miljöcertifierade fastigheter.
Förord
Masterexamensarbetet är skrivet under vårterminen 2020 och är slutgiltiga examinationen för civilingenjörsprogrammet Samhällsbyggnad, inom mastern Fastigheter och Byggande vid Kungliga Tekniska Högskolan.
Vi vill rikta ett stort tack till fastighetskonsultsbolaget Cushman & Wakefield, för den vägledning och information som vi har fått ta del av. Utan detta skulle arbetet inte varit möjligt att
genomföra.
Slutligen vill vi uttrycka vår tacksamhet till handledare Åke Gunnelin för stöd och värdefulla synpunkter.
Stockholm, maj 2020
Amanda Grafström & Hanna Käck
Table of Contents
1. Introduction ... 1
1.1. Background and context ... 1
1.2. Aim and research questions ... 2
1.3. Limitations ... 2
1.4. Outline ... 3
2. Literature Review ... 4
2.1. Environmental certifications ... 4
2.1.1. BREEAM... 4
2.1.2. GreenBuilding ... 5
2.1.3. LEED ... 5
2.1.4. Miljöbyggnad ... 6
2.1.5. Comparison of the certifications ... 6
2.2. Tenant demand ... 7
2.3. Transactions of environmentally certified properties ... 8
2.3.1. Price effects and assessed market values of sustainable buildings ... 9
2.3.2. Price effects of environmental certifications ... 10
3. Theoretical Framework ... 12
3.1. Theory of value ... 12
3.1.1. Value in use and value in exchange ... 12
3.1.2. Intrinsic value ... 12
3.1.3. Subjective theory of value ... 12
3.2. Marketing and branding theory ... 13
4. Research Method ... 14
4.1. Data collection ... 14
4.1.1. Possible shortcomings with the data ... 15
4.1.2. Variables in the regression analysis ... 16
4.2. Regression model ... 21
4.2.1. Hedonic pricing model & regression analysis ... 21
4.2.2. Propensity scores ... 23
5. Results ... 25
5.1. Model 1 ... 25
5.1.1. Linear regression ... 25
5.1.2. Linear regression with propensity scores ... 27
5.2. Model 2 ... 29
5.2.1. Linear regression ... 29
5.2.2. Linear regression with propensity scores ... 29
5.3. Model 3 ... 30
5.3.1. Linear regression ... 30
5.3.2. Linear regression with propensity scores ... 30
6. Discussion ... 31
7. Conclusions ... 34 References
Appendix
1
1. Introduction
1.1. Background and context
During 2017, the greenhouse gas emissions from the construction and real estate sector amounted to 19% of the total emissions in Sweden1 (Boverket, 2020a). Out of these 19%, the construction industry contributed to 50% of the emissions and property management constituted the other half.
Compared to 2016, the emissions are increasing and the energy consumption of the construction and real estate sector has risen since 2008 (Boverket, 2020b). In 2045, at the latest, Sweden shall have zero net emissions of greenhouse gases. To be able to reach the goal, the government will institute a requirement for newly produced buildings to state and declare their greenhouse emissions. As a help, there are a number of environmental certification systems which assess the performance of the buildings, from a social, environmental and economical perspective (Sweden Green Building Council, 2020a). There are now more environmentally certified properties than ever and some of the commonly used certifications in Sweden are GreenBuilding, Miljöbyggnad, LEED and BREEAM.
Overall, developing an environmentally certified property is known to entail additional costs during the development phase (Fuerst & McAllister, 2011a). For a property developer, the relationship between the price and the production cost is of great interest. As a compensation for the high additional costs, the developer and owner obtain benefits such as reduced operational costs and, according to BREEAM (n.d., b) and LEED (U.S. Green Building Council, 2020a), a higher asset value of the property. A common driver for property owners to aim for
environmental certifications is the belief of receiving a higher property value, as a result of the certification (Falkenbach, Lindholm & Schleich, 2010). Therefore, it is of interest to investigate whether the statement of BREEAM (n.d., b) and LEED (U.S. Green Building Council, 2020a) is true: do environmental certifications generate higher values of properties? Further, is the high value a result of lowered costs or does the certification in itself generate a higher price?
Previous research has shown there do exist a price premium (e.g. Fuerst & McAllister, 2011a;
Chegut, Eichholtz & Kok, 2014; Fuerst, 2009; McGrath, 2013; Robinson & McAllister, 2015;
Holtermans & Kok, 2017) for properties which are classified with an environmental certification.
However, the majority of the conducted studies are based on properties which are located in the U.S. or on the Asian market. Very few have been undertaken in Europe, especially in the Nordic region. Another lack in previous studies is that the majority are only examining the effect of LEED certifications (Falkenbach, Lindholm & Schleich, 2010) and the internationally used
1 Seen from a life cycle perspective
EnergyStar. It is therefore of interest to examine if the price premium does as well exist on the Swedish property market, with other certifications than solely LEED and EnergyStar.
In this study, a regression analysis was performed to investigate the impact environmental certifications have on office transaction prices. The regression was carried out on a main model which, among other parameters, accounted for whether the property was environmentally certified or not. Apart from the main model, two additional models were developed. The second model investigated whether the property had a certification of group A (GreenBuilding and Miljöbyggnad) or group B (BREEAM and LEED) and the third model examined the impact of each environmental certification. The used data in the analysis was mainly obtained from Cushman & Wakefield.
1.2. Aim and research questions
The aim of the thesis is to explore if there exists a price premium for office properties, which are classified with any environmental certification and are located in Sweden. Further, the aim is to investigate whether the potential price premium is a result of lowered costs or if the
environmental certification in itself generates a higher value of the property. The study will contribute with knowledge regarding whether other environmental certifications than LEED and EnergyStar affect the property value and generate a price premium. In addition, the research will conduce with information regarding what the price premium is a consequence of. Overall, the study will expand the existing research regarding price premiums of environmental certifications on the European market.
The research questions which will be answered are the following:
- Does there exist a price premium for environmentally certified office properties on the Swedish market?
- If there exists a price premium, is the premium a result of reduced cost levels of the property, or does the branding of possessing a property with an environmental certification generate a higher price?
1.3. Limitations
The data collection and analysis have been limited to transactions undertaken on the Swedish property market, more specifically in the three largest cities of Sweden: Stockholm (the entire region), Gothenburg (including Mölndal) and Malmö. The studied segment is office properties and selected time period is transactions which were completed as from 2010 through the first quarter of 2020. Transactions where only a share of the property was sold have been removed from the data set. Likewise for portfolio transactions which contain a larger number of properties.
3 The study has been limited to solely examine the effect of the following certifications: LEED, GreenBuilding, BREEAM and Miljöbyggnad.
1.4. Outline
Chapter 2, Literature Review:
A summarisation of relevant research is presented, containing information regarding each
environmental certification and its requirements. In addition, the certifications are compared with each other. Further, research of current demand from tenants regarding green buildings and environmental aspects is presented as well as previous studies on price effects of sustainable buildings and environmentally certified properties.
Chapter 3, Theoretical Framework:
Relevant theories for the research question are presented whose purpose is to act as a foundation when explaining and discussing the research result. The described theories are different value theories, marketing and branding theories.
Chapter 4, Research Method:
The used methods in the thesis are described by the data collection process and explanations of the regression analysis as well as the used variables are presented. In addition, discussions of potential shortcomings with the gathering of the data and the chosen analysis is displayed.
Chapter 5, Results:
The results from the log-linear regression are displayed as well as the log-linear regression with propensity scores. The presented result consists of the main model, as well as the second and third model.
Chapter 6, Discussion:
A discussion is presented regarding the results of the different regression analyses and its comparison to previously conducted studies. The result is further compared to the theoretical framework, and potential shortcomings of the outcome is as well displayed.
Chapter 7, Conclusions:
The conclusion contains an outline of the discussion regarding conclusions drawn from the result.
2. Literature Review
2.1. Environmental certifications
Environmental certifications contribute to a more sustainable development (Sweden Green Building Council, n.d.). The current most commonly used certifications in Sweden are
BREEAM, LEED, GreenBuilding and Miljöbyggnad (Swedish Concrete Association, 2013). In Sweden, 415 office properties have been certified with these environmental certifications during the years 2010 and 2020 2 (Sweden Green Building Council, 2020a). Another common
certification is Svanen. The system can however solely be applied on small houses, multi- apartment dwellings, schools and preschools (Svanen, 2020). Therefore, the study has not included this certification since it does not comprise office properties.
2.1.1. BREEAM
The international certification system BREEAM, short for Building Research Establishment Environmental Assessment Method, is one approach to assess how sustainable a building or project is (BREEAM, n.d., a). A licensed assessor evaluates the performance of the real estate to ensure the property or project is achieving the different requirements within the BREEAM certification program. The environmental certification was established in 1990, United Kingdom and more than 500,000 properties around the world have received the certification (Sweden Green Building Council, 2018). According to BREEAM (n.d., b), there are several reasons to certify a building, one of which is that properties with sustainable characteristics are sold with a price premium. To support the statement, BREEAM is referring to a study published by the Royal Institution of Chartered Surveyors (RICS) whose result indicated a 21% price premium for office buildings, certified with BREEAM and located in London.
BREEAM has several different technical standards and for an already developed office building situated in Sweden, the most convenient standard according to BREEAM (n.d., c) is BREEAM international in-use. The standard is considering following environmental categories: “...Energy, Water, Transport, Management, Waste, Pollution, Health & Well-Being, Land Use & Ecology and Materials” (BREEAM, n.d., d). There exists a Swedish version of BREEAM called BREEAM-SE which is adjusted and developed in collaboration with Sweden Green Building Council. The system is used to certify newly constructed buildings (Sweden Green Building Council, 2018) and has made it possible to apply an international certification system in
accordance with Swedish regulations and standards (Stadsbyggnadsnämnden, 2019). In addition, BREEAM-SE enables receiving the certification in Swedish. The property is assessed depending on the amount of achieved points in each category. The total point is aggregated to a certain grading, of which there are six levels: Unclassified, Pass, Good, Very Good, Excellent and
5 Outstanding, whereof Outstanding is the toughest level to receive (Sweden Green Building
Council, 2018). Further, extra points can be given to a building which have obtained innovative technical solutions.
2.1.2. GreenBuilding
GreenBuilding was originally an initiative by the European Union which took place during 2004 to 2014 (Sweden Green Building Council, 2019a). The aim of creating the certification was to improve the energy efficiency in the real estate and construction sector. Since 2010, Sweden Green Building Council is responsible for the operation in Sweden. To obtain the certification, the property shall reduce its energy consumption by 25% alternatively, if newly constructed, have an energy consumption which is 25% less than what is required for new developments. There are also other elementary requirements which need to be achieved (Stoll, 2019a; Stoll, 2019b). For instance, an energy management system must be installed in the property, which applies to both older and newly produced buildings. In addition, to be able to show fulfilments of the energy requirements, descriptions, calculations and/or measures must be stated clearly.
Regardless if the property is reconstructed or under construction, the GreenBuilding certification can be achieved (Sweden Concrete Association, 2013). A property owner has two different partnership alternatives. One option is to receive the certification at a property level, meaning one or several properties receive the certification. The second alternative is to be a corporate partner, resulting in requirements of the entire real estate portfolio. These requirements state that 30% of the already developed properties must be certified and 75% of the new developments. Today, there are more than 360 properties certified with GreenBuilding in Sweden (Sweden Green Building Council, 2020a).
2.1.3. LEED
The LEED certification, short for Leadership in Energy and Environmental Design, is the most widely adopted environmental certification around the world (U.S. Green Building Council, 2020a). The system can be applied on all kinds of properties. Similar to the BREEAM certification system, properties with LEED obtain various ratings depending on the achieved points (Sweden Green Building Council, 2020a). The grading system is divided into four
different levels: Certified, Silver, Gold and Platinum, of which Platinum is the hardest to achieve.
The points are given in regard to six main classes which are Innovation, Materials and Resources, Energy and Atmosphere, Water efficiency, Sustainable sites and Indoor environmental quality (King Sturge, 2009). Further, LEED argues there are economic benefits when classifying a property with the certification (U.S. Green Building Council, 2020a). Those are lower vacancy rates, higher rental income and, most importantly, an increased asset value. As for BREEAM, LEED does as well have different rating systems depending on if the property is a new
construction or during major renovation and if the interior space is up for redesign (U.S. Green Building Council, 2020c). Other examples of different systems are those adjusted for building
operations and maintenance or for neighbourhood development. A disadvantage of the certification is its lack of local adjustments to the Swedish market (Sweden Green Building Council, 2020c).
2.1.4. Miljöbyggnad
Miljöbyggnad is a Swedish environmental certification and there are over 1,500 properties which have received the classification (Sweden Green Building Council, 2020b). Both developed properties and new constructions can be certified and graded based on a three-level system:
Bronze, Silver and Gold (Sweden Green Building Council, 2019c.). To obtain the latest version, Miljöbyggnad 3.0, the property must fulfil 15 requirements. A few examples are thermal power, solar heat load, share of renewable energy, noise, radon and ventilation. The assessment mainly constitutes of measuring the air quality and the levels of nitrogen dioxide as well as radon (Sweden Green Building Council, 2020b). In addition, the daylight is considered to be important and is therefore measured to ensure a stable level. Within three years as from the building is certified, a check-up is performed to guarantee the building still fulfils the requirements of the certification.
2.1.5. Comparison of the certifications
Below, the various criteria for the environmental certifications are presented in Table 1, in order to distinguish the certifications. Similar comparisons, shown in Table 1 and 2, have been made by for example Reed et al (2009). In Table 1, the conditions have been simplified and some aspects have been sorted to the same type of criteria, see Appendix 1. For instance, indoor climate involves solar heat load, daylight, thermal climate, noise and ventilation. The table has been created to get an overview of each system and is not an in-depth comparison study between the certifications. As seen in the table, LEED and BREEAM are taking into account the largest number of criteria. LEED considers all areas shown in the table. In contrast to Miljöbyggnad, LEED and BREEAM, GreenBuilding only includes the energy aspect. Likewise, the Swedish Concrete Association (2013) claims that GreenBuilding solely considers the energy aspect while LEED, Miljöbyggnad and BREEAM include additional parameters, for instance material.
Moreover, BREEAM and LEED are as well considering variables such as the surrounding areas.
GreenBuilding is a cheaper system which is easy to understand. On the other hand, the energy measurements have to be made very accurately. LEED and BREEAM are internationally known and might be of interest to the international market (Swedish Concrete Association, 2013).
Further, an advantage with BREEAM is the adjusted system which is adapted to Swedish construction rules, in contrast to LEED. In addition, BREEAM, LEED and Miljöbyggnad
consider more aspects which could be a disadvantage, since there are more conditions to account for.
7
Table 1 Comparison of different criteria for each certification
Criteria BREEAM GreenBuilding LEED Miljöbyggnad
Energy X X X X
Health & well-being X X X
Indoor climate X X
Innovation X X
Local environmental
conditions X X
Management X X X
Material X X X
Pollution X X
Transport X X
Waste X X
Water X X X
Source: BREEAM, n.d., d; King Sturge, 2009; Sweden Green Building Council, 2019a; Sweden Green Building Council, 2019c;
Sweden Green Building Council, 2018; Vattenfall n.d.
Almost all of the different certifications contain various levels, of which some are easier to achieve and others harder. In Table 2, a compilation is presented of each certification and its different achievable levels. BREEAM is the environmental certification which has the highest number of levels. GreenBuilding on the other hand does not have any levels, a property can only be certified with GreenBuilding or not certified.
Table 2 Compilation of the different certification levels
BREEAM GreenBuilding LEED Miljöbyggnad
Unclassified Certified Bronze
Pass Silver Silver
Good Gold Gold
Very Good Platinum
Excellent Outstanding
Source: Sweden Green Building Council, 2019c; Sweden Green Building Council, 2020a; Sweden Green Building Council, 2018
2.2. Tenant demand
When tenants are considering new office premises, there are a number of factors which influence their decision. Some important aspects are location, cost (Kim, Lim & Kim, 2019) staff
requirements, marketing and availability (Levy & Peterson, 2013). Tenants have as well started to consider sustainability as a more important parameter when relocating to a new office. How important each of these parameters are depends on the size and type of the organisation. Clearly, the need to locate in an environmentally certified or sustainable building is not the only
influential factor when allocating. Depending on if the organisation is large or small, the
sustainability concerns vary. A larger organisation is more concerned about whether the building satisfies the company’s Corporate Social Responsibility (CSR) policy, how it affects their
stakeholders and if the perception of the company would change (Collins, Junghans & Haugen, 2016; Levy & Peterson, 2013; Livingstone & Ferm, 2017). Smaller businesses are more
constrained in terms of costs and if they choose to allocate in a sustainable building, the decision maker is often passionate about sustainable practices and is implementing operational energy efficiency policies (Levy & Peterson, 2013).
Renting an office space in a green building is seen to have a symbolic as well as a tangible value for the tenants (Kim, Lim & Kim, 2019). Green buildings are considered to be something
marketable for the tenant and can appeal to the public as well as clients. For larger organisations, to locate in a sustainable building has become the market norm (Levy & Peterson, 2013). This does fulfil company policies as well as reflect an environmentally conscious image. To several tenants, sustainability is linked to the perception of the market, instead of a scientific approach on carbon neutrality. There exists a willingness to pay extra for leases which are signed in green buildings (Robinson et al, 2016; Kim, Lim & Kim, 2019). Bond & Devine (2016) even state that all properties which claim to be green have a rental premium. How large this premium is varies depending on the conducted study, from 3% to 19.7% (Eichholtz, Kok & Quigley, 2013; Fuerst
& McAllister, 2011a; Robinson et al, 2016; Chegut, Eichholtz & Kok 2014; Holtermans & Kok, 2017; Eichholtz, Kok, & Quigley, 2010; Fuerst & McAllister, 2011b). Chegut, Eichholtz & Kok (2014), whose result was 19.7%, performed a transaction-based hedonic model with a sample of BREEAM-rated office buildings in London. Eichholtz, Kok & Quigley (2013) did undertake a similar analysis on LEED and EnergyStar rated office buildings in the U.S., but instead
concluded the rent premium of green properties to be 3%. The rental premium seems to add a higher value in regions and smaller markets, as well as in peripheral parts of larger cities where the overall rent is lower (Eichholtz, Kok, & Quigley, 2010).
2.3. Transactions of environmentally certified properties
There are a number of reasons for property owners to achieve environmental certifications on real estate (Lind & Lundström, 2009). These are for example better branding, increasing rents, lower vacancy rates, decreased maintenance and operation costs as well as a receiving a higher market value. Brown (2017) states that the real estate owners believe environmentally improved and certified buildings have higher market values as well as increased production costs. The aim is to undertake the most attractive property as possible, given the price, and sustainability is a key factor. Cadman (2000) has developed a sustainability model, shown in Figure 1, in which different stakeholders on the market and their relationships have been identified. The model could be seen as outdated, since it demonstrates why there is a lack of sustainability in the sector which is not a larger issue nowadays. However, the model is still essential due to its display of the crucial actors on the market. If there is no unified view, the sustainable development will cease and there will be no equilibria of demand and supply.
9
Figure 1 The vicious circle of blame (Cadman, 2000)
A large problem with sustainable properties is the lack of knowledge and understanding of their potential benefits (Christensen, 2017). Investors do actively look for opportunities of which they can minimise risk, improve their competitiveness on the market and create value with sufficient property management in their portfolios. However, if the brokers do not understand sustainable properties, a barrier is created which mitigates the investments. By contrast, for high-end commercial properties located in some markets, environmental certifications have become expected. As a result, the certifications are no longer a factor which can differentiate the
properties from their competition. When broadening the perspective to all markets and qualities, company owned properties are more likely to become environmentally certified, compared to real estate owned by private investors (Qiu, Su & Wang, 2017). Several property owners, which are market leaders, prefer to use international environmental certifications such as LEED and BREEAM (Brown, 2017). This is because the certifications are known from an international perspective, in which a value chain will be created with a global trend as a result (Lind &
Lundström, 2009).
2.3.1. Price effects and assessed market values of sustainable buildings
Some of the previously conducted studies within this field have investigated if more sustainable buildings, so called green buildings, have a price premium or a rental premium compared to non- or less sustainable buildings (Bond & Devine, 2016; Davis et al, 2015; Cerin, Hassel &
Semenova, 2014; Chi-Man Hui, Tse & Yu, 2014; Popescua et al, 2012; Eichholtz, Kok, &
Quigley, 2010). In other words, these studies are not examining whether the environmental
Occupiers
"We would like to have sustainable buildings but
there are very few available"
Constructors
"We can build sustainable buildings, but the developers don't
ask for them"
Developers
"We would ask for sustainable buildings, but the investors won't
pay for them"
Investors
"We would fund sustainable buildings, but there is no demand
for them"
certification generates a price premium. Instead they are addressing properties which claim to be green and sustainable. The greenness of a building is becoming more and more significant (Fuerst, 2009) and the sustainability impact in the real estate industry is heavy (Abdullah, Mohd
& Sabu, 2016). In general, it has been shown that properties which are claiming to be sustainable and green have a rental premium (Bond & Devine, 2016). There also seems to be a connection between energy efficiency and sales price, the more energy efficient the building is, the higher the price (Davis et al, 2015; Cerin, Hassel & Semenova, 2014; Popescua et al, 2012; Eichholtz, Kok, & Quigley, 2010).
Furthermore, previous research discuss how property valuers are addressing sustainability when assessing the market value (Warren-Myers, 2012; Warren Myers & Reed, 2010; Lorenz &
Lützkendorf, 2011; Rahman, Rowlands & Weber, 2017). In many cases, if the building is green or not is more or less not taken into consideration by the valuer. This is because there exists a lack of knowledge as well as an understanding of sustainability (Warren-Myers, 2012; Warren- Myers, 2015) and the elements which contribute to the sustainability of a building (Warren- Myers & Reed, 2010), especially among younger valuers (Warren-Myers, 2015). However, energy efficiency seems to have a very strong positive influence on the assessed market value (Warren-Myers, 2013). Students are recommended, as well as those already working in the valuation profession, to be taught how sustainability affects the value of the properties (Warren- Myers, 2015). If there is a greater knowledge within the real estate sector regarding sustainability, the sector will move towards a sustainable future. Previously, the lack of knowledge within the sustainability field was even greater and the adoption of sustainability was limited in the real estate market. Today, the knowledge is more widespread although the appraisal industry does not seem to estimate a higher value for properties with environmental certifications compared to non- certified (Rahman, Rowlands & Weber, 2017).
2.3.2. Price effects of environmental certifications
In addition to the research of sustainable and green buildings, several studies have investigated the price difference between environmentally certified buildings and non-certified buildings (Fuerst & McAllister, 2011a; Chegut, Eichholtz & Kok, 2014; Addae-Dapaah & Chieh, 2011;
Popescua et al., 2012; Abdullah, Mohd & Sabu, 2016; Dermisi, 2009; Robinson & McAllister, 2015; Holtermans & Kok, 2017). A majority of the research is performed with a regression analysis, in which the price is the dependent variable. The studies have shown there do exist a price premium for certified office buildings compared to non-certified (Fuerst & McAllister, 2011a; Fuerst & McAllister 2011b; Chegut, Eichholtz & Kok, 2014; Addae-Dapahh & Chieh, 2011; Fuerst, 2009; McGrath, 2013; Abdullah, Mohd & Sabu, 2016; Robinson & McAllister, 2015; Holtermans & Kok, 2017; Mangialardo, Micelli, & Saccani, 2019; Fuerst & McAllister, 2011b; Eichholtz, Kok & Quigley, 2013; Eichholtz, Kok & Quigley, 2010; Miller, Spivey &
Florance, 2008). The scope of the premium is depending on the market and the type of certification. A majority of the research was performed on the U.S. market, comprising the
11 environmental certifications LEED and EnergyStar. How large the price premium is varies
heavily between the conducted studies. Some researchers did solely perform regression models which examined whether office properties with any environmental certification were sold at a higher price. Their result fluctuated between a price premium of 10.1% to 16.0% (Holtermans &
Kok, 2017; Eichholtz, Kok & Quigley, 2013; Eichholtz, Kok & Quigley, 2010). Others
performed regression models to determine if specific certifications affected the premium more or less than others. The price premium of EnergyStar-certified office properties varied between 5.8% to 26.0% and for LEED 9.9% to 32.0% (Miller, Spivey & Florance, 2008; Fuerst &
McAllister, 2011a; Fuerst & McAllister, 2011b; Robinson & McAllister, 2015). Mere one found study has performed a regression analysis on the European market. Chegut, Eichholtz & Kok (2014) did research BREEAM-certified properties in London and received a result of which certified properties was sold at a price premium of 14.7% compared to non-certified properties.
Nevertheless, the size of the office property does matter. Overall, smaller buildings tend to possess premiums, while there is limited proof of a premium in larger office buildings (Robinson
& McAllister, 2015). Generally, there is a higher return because of lowered operating costs for office buildings which are certified, and the price premium seems to be larger if the property is classified at the highest level (Mangialardo, Micelli, & Saccani, 2019). Similar results have been found in a Singaporean housing market study (Deng, Li & Quigley, 2012). However,
Fesselmeyer (2018) did as well conduct a housing study on the Singaporean market and found an opposite result: the premium is larger for properties which are certified at the lowest level. The author does not clearly motivate why, but states that this occurs for developments whose
greenness is less distinct to the buyer, and for whom the certification in itself is more important.
The demand from tenants is an important driver for property owners to obtain one or several environmental certifications on their properties (Brown, Malmqvist & Wintzell, 2016). There are numerous researchers who have concluded that property owners are using environmental
certifications in order to improve their marketability, instead of amending their performance (Matisoff, Noonan, & Mazzolini, 2014; Reichardt, 2014; Freybote & Qian, 2017; Qiu, Su &
Wang, 2017; Christensen, 2017). A reason for a retailer to obtain an environmental certification in their store is the intangible effects such as reputational benefits (Reichardt, 2014; Freybote &
Qian, 2017). Retailers are mainly motivated by reputational benefits rather than achieving tangible effects such as lower operating costs (Freybote & Qian, 2017). The rent premium which comes with a green building has been found to be greater than the cost savings (Qiu, Su & Wang, 2017).
3. Theoretical Framework
3.1. Theory of value
Economic value, at a given time, is seen as the monetary worth of a property, goods or service to sellers and buyers (Mooya, 2016). A value is intangible but there is a need for a quantitative measurement in economics. This measurement is money, which could in a broader perspective be expressed as resources paid to acquire a good. The theory of value can be described from several different economic perspectives, examples are listed below.
3.1.1. Value in use and value in exchange
The concept of customer value can be divided into two different perspectives: value in use and value in exchange (Eggert et al, 2018). The theory of value was mainly focused on value in exchange until conceptualisations evolved and changed perspective to a value in use.
Before, the focus was on resources as well as value in exchange. Now, the emphasis is on resource integration and value in use. Changing the perspective has developed a new and fresh view of customer value proposition. In the real estate sector, value-in-use is seen as the market value for a particular user (Gadd, 1989). This concept can be applied on the research since the value to an individual can vary, the importance of sustainability is for example not equal to everyone. Therefore, when conducting the research it is crucial to be aware that the worth of a property to an individual is not always equal to the market value. Though, this is hard to measure.
However, knowing the concept of individual worth and understanding its impact when real estate transactions are undertaken is important.
3.1.2. Intrinsic value
The intrinsic theory of value takes into consideration the calculated or perceived value of an asset, investment or company (Kenton, 2019). The value is characterised as a value which something has “in its own right”, “as such”, “for its own sake” or “in itself” (Zimmerman &
Bradley, 2019). A key meaning of the theory is the difference between valuable for its own sake as opposed to valuable for the sake of something else. A crucial part in this value is moral
judgement as well as decisions, since the value is based on the attitude of the valuer. An intrinsic value does consider trademarks, brand name as well as copyrights which normally is not
calculated or accurately reflected in the market price. An intrinsic value does not reflect the investor’s willingness to pay, instead it indicates what the object is really worth.
3.1.3. Subjective theory of value
According to the subjective theory of value, the worth of an object is different depending on how much a person desires or needs the object in question (Kagan, 2018). Instead of basing the value of the object on hours of labour or resources, the theory places value on how useful and scarce
13 the item is. Furthermore, an entailed shortcoming of the theory is the problem to measure a value consistently. Value can be created or increased by simply transferring the ownership to a person who regards the object at a higher value. The value of an object is influenced by cultural
significance, sentimentalism, situational circumstances as well as nostalgia. The value is assessed as the willingness to pay and is displayed when objects are put up for auction. During a sale, there is an indication of the value the bidders believe the object holds.
3.2. Marketing and branding theory
From a marketing perspective, companies, property owners and tenants might be more or less willing to pay for an environmentally certified building. The kind of aspects which affect the client's perception of an actor is further described. “A brand is a name, term, design, symbol or any other feature that identifies one seller’s good or service as distinct from those of other sellers” (American Marketing Association, n.d.). The brand is seen as a much more complex symbol than one where the seller merely is delivering the services, characteristics and advantages (Kotler, 2000). However, the meaning of a brand can consist of attributes, benefits, values, culture, personality and the user. An example of an attribute of a real estate company could be possessing high-quality properties. These attributes need to be transformed to some sort of benefit. For instance, one could say: if high quality buildings are purchased or owned, there is no need to renovate the building in the near future. As mentioned, values and culture do as well affect the brand (Kotler, 2000). For example by incorporating environmental awareness and a corporate culture of openness. Regarding personality, the brand can be represented by a certain person, animal or object according to Kotler (2000). Lastly, a user is reflected by the brand due to the public associating a certain type of user to a specific brand. Out of all the components, value is one of the three elements of a brand which are more persistent.
To be able to produce a successful product or offering, it is important to bring value and
gratification. The value to the customer depends on “...what the customer gets and what he gives”
according to Kotler (2000), whereof the benefits are what the customer receives, and the costs are what the client gives. The benefits can be divided into emotional and functional benefits. The emotional benefits are defined as “...something nice I can say about myself because I use your product or service.” (Livingston, 2010) while the functional benefits are described as the benefits which could be linked to the act of the certain product or service (Zorraquino, n.d.). To raise the value, the marketer can increase benefits, lower costs or combine the two options (Kotler, 2000).
4. Research Method
The study has been conducted with a quantitative approach (Saunders, Lewis & Thornhill, 2016).
When using this procedure, it is significant to have a highly structured methodology. This is because of the necessity to make the study replicable and ensure its reliability. Another important characteristic with the approach is the need for generalisation. The sample must be large enough and has to be selected carefully. Additionally, in order to ensure the validity of the inquiry the measures and the study of the result need to be considered.
The quantitative approach enables one to explore relationships between variables within the data (Saunders, Lewis & Thornhill, 2016). To examine statistical relationships, different software packages can be used. In this study, Stata was the chosen statistical software. As earlier mentioned, similar studies have used a hedonic regression model to be able to examine if the value of a property is higher for an environmentally certified building. Accordingly, a hedonic regression model has been used. The main regression model in the analysis, further referred to as Model 1 or the main model, considered whether the property was environmentally certified or not. Two extra models, similar to Model 1, were as well developed and performed. Model 2 examined if the property was certified with a certification in group A (GreenBuilding and Miljöbyggnad) or B (BREEAM and LEED), further motivated and described in section 4.1.2., while Model 3 investigated each environmental certification and their potential impact:
Miljöbyggnad, GreenBuilding, BREEAM and LEED.
4.1. Data collection
All used data is secondary, compiled data which have received some sort of summarisation or selection. Since this thesis is in collaboration with Cushman & Wakefield, the majority of the data has been collected by the company. The information provided from Cushman & Wakefield consists of 343 office property transactions undertaken between 2010 and the first quarter of 2020, in which 56 were environmentally certified and 287 not certified. There were 18 certified properties in group A and 40 certified properties in group B. 12 properties were certified with Miljöbyggnad, 7 GreenBuilding, 19 BREEAM and 21 LEED. The data consists of information regarding the transaction price, transaction year, building area, costs as well as rental income.
Additional information connected to the transactions have been gathered by the authors, further explained in section 4.1.2. A disadvantage with secondary data is that the data might have been collected with a purpose in mind which is not equal to what is needed (Saunders, Lewis &
Thornhill, 2016). Another con is the lack of control regarding the data quality, all data sources must be carefully evaluated. Some caution must be undertaken when gathering secondary sources since they at first can appear as relevant, but at a closer look are not suitable for the research question.
15 Rumsey (2015) claims that most data is either categorical or numerical, whereof categorical data is referred to as qualitative data while numerical is allocated to quantitative. Further, both types of data can include variables with numerical values, yet the values of the categorical data do not have a meaning. In addition, categorical data can answer yes or no questions. In contrast to the categorical data, the numerical is represented by numerical values which can be measured, such as price of the property, rental income and cost. The used data is of categorical and numerical characteristics. A dummy variable can be defined as an indicator variable which is based on categorical characteristics (Hill, Griffiths & Lim, 2012). Commonly, the numbers 1 and 0 are used to determine whether the attribute is true or false. This method enables one to transform a non-numeric variable into a numerical. Hence, dummy variables have been used to convert the categorical variables.
4.1.1. Possible shortcomings with the data
When collecting the data, Creswell (2014) advocates the random sampling method. The approach means there is an equal probability for the entire population to be chosen. The population consists of all existing office properties, both sold and not sold. Thereof, the used data in the analysis solely represents a part of the population. Accordingly, it is crucial to study the concept of selection bias. Selection bias is often related to a non-random selection, consequently the study and the control groups are not comparable (Institute for Work & Health, 2014). The phenomenon can be described as follows: some products and assets are by the market considered to be more desirable, because of their value, and will be traded more regularly (Adam, 2013). To address the selection bias problem, propensity scores have been used, further explained in section 4.2.2.
The opted data collection method defines the experiment as either controlled or uncontrolled (Hill, Griffiths & Lim, 2012). A controlled experiment occurs when the purpose of gathering data initially is to carry out an experiment. An uncontrolled experiment on the other hand takes place when the data, at first, is collected for another reason. In this analysis, the majority of the data was gathered for another purpose. In a controlled study, it is possible to determine the effect of each independent variable, in contrast to the uncontrolled experiments. These uncontrolled experiments may cause collinearity, meaning: the variables are systematically changing in the same direction (Hill, Griffiths & Lim, 2012). For example, if there is a perfect linear relationship between independent variables, such as in the equation x1=10 + 2x2, then there is an exact
collinearity between x1 and x2 (Kim, 2019). This causes difficulties when explaining relationships between the dependent and the independent variable (Hill, Griffiths & Lim, 2012). A concept related to collinearity is multicollinearity, when there are more than two independent variables within the relationship (Kim, 2019). If multicollinearity occurs, there is a risk of misinterpreting the results from the regression model. Therefore, in order to construct a reliable regression model, the multicollinear independent variables have to be excluded. This is especially a problem when using dummy variables, explained as “falling into the dummy variable trap” (Hill, Griffiths &
Lim, 2012). The dummy trap is a result of exact collinearity. To avoid this error, one indicator
variable is usually omitted. This variable will instead determine the reference group.
Furthermore, other potential problems could be the occurrence of heteroskedasticity and
autocorrelation. Heteroskedasticity is occurring if all variances of the observations are different from each other while autocorrelation is described as variables related to one another over time.
As similar to other studies, a problem is the occurrence of missing data. This is a common issue according to Rubin (1976). When the missing data does not depend on the observed data, it is called Missing Completely at Random, MCR (Dong & Peng, 2013). In this study, there are some missing cost-data which can be classified as MCR due to its complete randomness. To manage the problem, several approaches are suggested (Kang, 2013). One is to replace the unknown variable with a mean estimation of another parameter. This method might cause inconsistent bias and is often not acceptable when handling missing data. Another approach is regression
imputation, a method of which the missing data is replaced with equivalent possible gathered values. By using this method, there is no need to delete data and the observations can be kept in the analysis. As a result, problems which come with a change in the standard deviation and distribution are avoided. Thus, the data set is greater and the standard error is at a lower level.
Due to the mentioned benefits, the second method has been adapted in this analysis by replacing missing cost-values with equivalent possible values, described in section 4.1.2.
4.1.2. Variables in the regression analysis
Below, all variables in the regression analysis are listed including information regarding their definition and the collection process.
Price
Price is the dependent variable in the regression model. The parameter is expressed as SEK per square meter and thereof is a numerical variable. The price of each transaction has been gathered by Cushman & Wakefield.
Difference transaction year and value year
The standard and condition of a property will affect its value. The Swedish Tax Agency
(Skatteverket) assigns a value year to all properties which are not classified as non-taxable. The value year is defined as the new construction year, which is adjusted if the property has been redeveloped (Lantmäteriet, 2018). To be able to take into consideration the standards and conditions of the property when the transaction took place, this variable will be an indication displaying the difference between the transaction year and the value year. The smaller the difference, the better condition and standard of the buildings on the property. Information regarding the transaction year was provided by Cushman & Wakefield. The value year for each property, at the transaction date, was gathered from Datscha (2020).
17 Submarket
The submarkets are categorical variables and were therefore converted to dummy variables, one for each submarket. The submarkets are allocated in Stockholm, Gothenburg and Malmö, shown in Table 3. The purpose of the submarket variable is to group areas which are more or less attractive, in order to capture the various willingness to pay for properties located in different areas. The markets were classified and specified together with Cushman & Wakefield. The borders of each submarket are found in Appendix 2.
Table 3 Chosen submarkets in Stockholm, Gothenburg & Malmö
Stockholm Gothenburg Malmö
Stockholm CBD Gothenburg CBD Malmö Old CBD & Västra Hamnen
Stockholm Inner City Gothenburg Gårda & Heden Malmö New CBD & Hyllie Stockholm Suburb Gothenburg Älvstranden &
Lindholmen Malmö Other
Stockholm Other Gothenburg Other (including
Mölndal) -
Distance to
To further explain the location of the property, this variable determines the linear distance, in metres, between the property and a central point in each city. For properties located in
Stockholm, the distance was measured from Sergels torg. For those located in Gothenburg or Malmö, the central measurement point was the respective central station.
Year
The property market, as well as the macroeconomic state, have changed during the past ten years which most likely have had an impact on the transaction price. Thus, time dummies were used to capture the different states of the Swedish property market when the acquisitions took place. The time dummies were expressed annually and based on the transaction year, gathered by Cushman
& Wakefield.
Cost
The cost variable is numerical defined as the annual cost in SEK per sqm. In the data provided by Cushman & Wakefield, not all observations did have cost information. Thus, this is a case of missing data. For those properties which did not have any information regarding costs, template costs were applied. In total, 48% of the data did not have any cost information. The cost variable varies significantly between properties depending on for example number of tenants, the size of the property and the structure of the rental agreements. To simplify, a normal cost value was assumed which were applied to all properties that did not have any cost information. Operation costs were estimated to be SEK 170 per sqm, maintenance SEK 70 per sqm and administrational costs SEK 20 per sqm. Due to the uncertainties regarding this parameter, each model was
performed in two different regressions of which one includes costs and the other excludes the variable. As a result, six different regressions were conducted.
Rental Income
Rental income is a numerical variable defined as the annual income in SEK per sqm, as if the property was fully rented. It was decided, together with Cushman & Wakefield, to not take into consideration the vacancy of the properties because of a smaller impact on the price. The rental income, during the transaction year, was gathered from Cushman & Wakefield and
complemented with the reported taxable income which was collected from Datscha (2020).
Environmental certification
The variable environmental certification is categorical and was converted to dummy variables.
The regression analysis was performed with three different models, in which solely the environmental certification variable was altered. In all three models, the environmental certification is displayed as dummy variables. Properties which were classified with an
environmental certification at the transaction date or received a classification within a couple of months afterwards, have been accounted as environmentally certified. Some environmental certifications have various versions in which there are different requirements, for example BREEAM. To minimize the number of variables, all different BREEAM versions have been classified as one. Information regarding if the properties were environmentally certified or not was gathered from Sweden Green Building Council (2020a), BREEAM (n.d., e) and U.S. Green Building Council (2020b).
In Model 2, the different certifications are paired together in group A (Miljöbyggnad &
GreenBuilding) and group B (BREEAM & LEED). The reason for this pairing is because of BREEAM and LEED being international certifications which do have a wider spread globally and are considered by the property industry to be more equal and tough to receive. Miljöbyggnad and GreenBuilding are instead Swedish certifications which in the real estate sector are seen as less difficult to obtain.
Omitted variables
Previously, the problem with exact collinearity, so called dummy traps, was mentioned. To solve the issue some variables must be omitted. Accordingly, following have been used as the omitted variables: year 2010, Malmö other, not certified in the main model, group B in Model 2 and GreenBuilding in Model 3.
The explanation of each specific variable is listed in Table 4 and a summarisation with descriptive statistics is presented in Table 5.
19
Table 4 Explanation of all variables in the regression models
Variable in equation Definition
log_sek_per_sqm The logarithmic of the dependent variable price. Displayed
as SEK per sqm.
diff_year The difference between the transaction year and the given
value year of the property.
cost_sek_per_sqm The cost of the property displayed as SEK per sqm.
rental_income_sek_per_sqm The income of the property, displayed as SEK per sqm.
year_XXXX All year variables are time dummies, one for each year
when the transaction took place.
location_sthlm_cbd The submarket Stockholm CBD. A dummy variable.
location_sthlm_inner_city The submarket Stockholm Inner City. A dummy variable.
location_sthlm_suburb The submarket Stockholm Suburb. A dummy variable.
location_sthlm_other The submarket Stockholm Other. A dummy variable.
location_gbg_cbd The submarket Gothenburg CBD. A dummy variable.
location_gbg_garda_heden The submarket Gothenburg Gårda & Heden. A dummy variable.
location_gbg_alvstranden_lindholmen The submarket Gothenburg Älvstranden & Lindholmen. A dummy variable.
location_gbg_other The submarket Gothenburg Other. A dummy variable.
location_malmo_newcbd_hyllie The submarket Malmö New CBD & Hyllie. A dummy variable.
location_malmo_oldcbd_vastrahamnen The submarket Malmö Old CBD & Västra Hamnen. A dummy variable.
distance _to The distance from the property to a central location in the
city the property is situated in.
certified If the property is environmentally certified or not. A dummy
variable.
certified_a If the property is certified with an environmental
certification, which belongs to group A. A dummy variable.
miljobyggnad If the property is certified with Miljöbyggnad. A dummy
variable.
BREEAM If the property is certified with BREEAM. A dummy variable.
LEED If the property is certified with LEED. A dummy variable.
Below, Table 5 is presented containing summary statistics. There are in total 343 observations of each variable. As seen in the table, of all-time dummies, year 2012 and 2014 have the highest mean value, meaning these are the years which were very transaction intense. 2020 has the lowest mean, which is naturally due to containing data solely from the first quarter. Regarding the
location of the properties, most transactions were undertaken for properties located in Stockholm Inner City, while Gothenburg Gårda & Heden as well as Malmö New CBD & Hyllie had the lowest amount of acquisitions. The average difference between the transaction year and the assigned value year of the property is about 30 years, average rental income is SEK 2,080 per sqm and cost SEK 337 per sqm. Out of all variables connected to environmental certifications, certified has the highest mean because of its inclusivity of all certified properties. Miljöbyggnad is the most uncommon certification of the data.
Table 5 Summary statistics of all variables in the regression analysis
Variable No. of
Observations Mean Standard
Deviation Min Max
log_sek_per_sqm 343 10.3005 0.7382549 7.460136 12.09177
diff_year 343 30.37609 23.28758 0 91
cost_sek_per_sqm 343 337.3969 166.7034 66.55434 1244.564
rental_income_sek_per_sqm 343 2081.079 991.3446 4 5667
year_2011 343 0.1137026 0.3179135 0 1
year_2012 343 0.1282799 0.3348897 0 1
year_2013 343 0.0962099 0.2953097 0 1
year_2014 343 0.1311953 0.3381071 0 1
year_2015 343 0.0991254 0.2992668 0 1
year_2016 343 0.0758017 0.2650673 0 1
year_2017 343 0.0874636 0.2829259 0 1
year_2018 343 0.090379 0.2871429 0 1
year_2019 343 0.0728863 0.2603295 0 1
year_2020 343 0.0233236 0.1511497 0 1
location_sthlm_cbd 343 0.090379 0.2871429 0 1
location_sthlm_inner_city 343 0.2857143 0.4524139 0 1
location_sthlm_suburb 343 0.1982507 0.3992645 0 1
location_sthlm_other 343 0.1166181 0.3214335 0 1
location_gbg_cbd 343 0.0437318 0.2047965 0 1
location_gbg_garda_heden 343 0.0291545 0.1684853 0 1
location_gbg_alvstranden_lindholmen 343 0.0233236 0.1511497 0 1
location_gbg_other 343 0.0991254 0.2992668 0 1
location_malmo_newcbd_hyllie 343 0.0291545 0.1684853 0 1 location_malmo_oldcbd_vastrahamnen 343 0.03207 0.1764433 0 1
distance_to 343 4436.12 6418.581 154 59300
certified 343 0.1632653 0.3701475 0 1
certified_a 343 0.0524781 0.223315 0 1
miljobyggnad 343 0.0349854 0.1840113 0 1
breeam 343 0.0553936 0.2290811 0 1
leed 343 0.0612245 0.2400919 0 1
