Energy Efficiency Renovations;
Author: Luuk Rietjens
Student number: S1014001
E-mail address: l.rietjens@student.ru.nl
Masterthesis
Master: Spatial Planning (Planning, Land & Real Estate Development)
Organisation: Radboud University
First thesis supervisor: Sander Lenferink
Second thesis supervisor:
Extern thesis supervisor: Chris Huppertz
Graduation period: February – July 2019
PREFACE
The question within the energy transition is not whether we are going to achieve the goals
from the climate agreement, but how we are going to pay for this? Parties from the
climate agreement believe that the energy transition can be affordable for everyone if we
succeed to reduce the costs through scaling up and innovation and ensure that the
monthly costs of the loan that you take out for the renovation do not exceed the benefit
you book on the energy bill.
Municipalities play a crucial role in this. Together with residents and building owners, they
need to consider carefully what the best solution is for each neighbourhood, if houses are
no longer heated with the traditional central heating boiler. The basic idea behind this
strategy is that when the demand for energy efficiency materials is organized, producers
can standardize their products and scale up their production process. Eventually the
supply of insulation materials and sustainable heat systems will be drastically increased
against reduced prices.
A neighbourhood-oriented approach, no matter how energetic, cannot succeed without
the co-operation of residents. Previous projects show that owner-occupied homes do not
opt for a serial approach, with entire streets taking the same measures. Property owners
need customization: each owner decides individually whether and, if so, what measures
he takes.
The energy transition seems to be caught between two fires. How do you ensure
customization on the one hand and upscaling on the other?
Commercial enterprises have shown for many years that these two go well together. This
strategy is also known as market segmentation or targeting.
Although market segmentation mainly has a commercial perspective, this logic can also be
translated into policy objectives, in this case in relation to the energy transition.
This thesis examines the segments that can be distinguished within the private housing
stock with regard to the heat transition. This knowledge enables policymakers and
professionals to align strategic choices with the characteristics of a specific group.
I believe, this study has developed valuable new information, on a scale that we have not
encountered before. By finding a scientific basis for the existence of different groups
within the energy transition, I am sincerely convinced that by applying this knowledge into
practice, this research positively contributes to the implementation of the climate
agreement.
I should like to thank my thesis supervisor Sander Lenferink and internship supervisors
Chris Huppertz and Judith van Rijswick for their feedback and substantive contributions. I
also want to thank my girlfriend in particular for her support.
ABSTRACT
The energy transition is high on the political agenda. At the end of February 2018, the
Minister of Economic Affairs and Climate on behalf of the government explained the
approach to climate policy with a letter to the House of Representatives. The goal of the
government is to reduce greenhouse gas emissions by 49% in 2030 compared to 1990 (K.
Beckman, 2018)
This can only be achieved if the energy transition is affordable for everyone. Parties from
the climate agreement believe that the energy transition can be affordable for everyone if
we succeed to reduce the costs through scaling up and innovation. Regarding the first,
municipalities play an important role. Together with residents and building owners, they
need to consider carefully what the best solution is for each neighbourhood, if houses are
no longer heated with the traditional central heating boiler.
A neighbourhood-oriented approach, no matter how energetic, is not automatically
successful. Previous projects have shown that owner-occupied homes do not opt for a
serial approach, with entire streets taking the same measures. Homeowners need
customization. The neighbourhood-oriented approach seems to be caught between two
fires; customization on the one hand and upscaling on the other.
Market segmentation is the key to success according to this research. By defining different
groups of homeowners within the energy transition, this research enables policy makers
and professionals to formulate strategies that match the characteristics of a specific
group. Ultimately this will lead to higher adoption rates.
OUTLINE
01
INTRODUCTION TO THE RESEARCH
9
01.01
The context of the problem
10
01.02
Research aim and research questions
11
01.03
Defining Energy efficiency Renovations
12
01.04
Defining the Dutch Home owners
13
01.05
Scientific and societal relevance of the research
14
01.06
Outline
15
02
LITERATURE REVIEW
17
02.01
Motivation concepts
17
02.02
Market segmentation
19
02.03
Likely motivated actors
20
02.04
Motivations towards EERs
21
02.05
Conclusion on the literature
23
03
METHODOLOGY
25
03.01
Philosophical viewpoint
25
03.02
Research strategy.
25
03.03
Conceptual model
27
03.04
Hypothesis
29
03.05
Research method
30
03.06
Method selection
33
04
RESULTS
38
04.01
WoON 2018 database
38
04.02
Regression analysis
41
04.03
Housing density
52
04.05
Relation between drivers and barriers and personal and contextual
factors. 56
04.06
Non-renovators
63
04.07
Discussion of result of statistical and regression analysis
70
05
CONCLUSION AND POLICY IMPLICATIONS
73
05.01
General recommendations
74
05.02
Recommendations (Renovators)
74
05.03
Recommendations (Non-renovators)
75
05.04
Limitations
77
05.05
Recommendations for further research
77
05.06
Reflection
78
06
REFERENCES
79
APPENDIX A CORRELATION MATRIX
84
APPENDIX B VIF STATISTICS
88
APPENDIX C BINARY LOGISTIC REGRESSION
90
APPENDIX D MULTI NOMINAL LOGISTIC REGRESSION (RENOVATORS)
97
APPENDIX E MULTI NOMINAL LOGISTIC REGRESSION (NON-RENOVATORS) 117
List of tables
Table 1 list of indicators found in literature. 25
Table 2 Selection of variables based on the literature and Housing survey Fout! Bladwijzer niet gedefinieerd.
Table 3 Response WoON2018 30
Table 4 Respond category linked to barrier 32
Table 5 Response category linked to driver Fout! Bladwijzer niet gedefinieerd.
Table 6 Casewise List 36
Table 7 Type of renovation 39
Table 8 percentage of household replacing the heating system and insulation of walls, floors or roof 39
Table 9 motive by type of renovation 40
Table 10 Example of Chi-square test for Age. 42
Table 11 Chi- square test EERs and Age. 43
Table 12 Cramer's V 43
Table 13 Adjusted standardized residual for age 44
Table 14 Chi-square test 45
Table 15 Logistic regression predicting likelihood of EERs on Age. 45
Table 16 Logistic regression predicting likelihood of EERs on Ethnicity 46
Table 17 Logistic regression predicting likelihood of EERs on Education 47
Table 18 Logistic regression predicting likelihood of EERs on Income 48
Table 19 Logistic regression predicting likelihood of EERs on building age 50
Table 20 Logistic regression predicting likelihood of EERs on type of household Fout! Bladwijzer niet gedefinieerd.
Table 21 Logistic regression predicting likelihood of EERs on type of house 51
Table 22 Logistic regression predicting likelihood of EERs on level of social cohesion 53 Table 23 Logistic regression predicting likelihood of EERs on perceived level of maintenance of the neighbourhood
54 Table 24 Logistic regression predicting likelihood of EERs on beliefs about energy efficiency Fout! Bladwijzer niet gedefinieerd.
Table 25 Difference between social demographic indicators and EERs 54
Table 26 Goodness of Fit 57
Table 27 Likelihood Ratio Test 57
Table 28 Social demographic characteristics of segments 62
Table 29 Number of respondents who reported that their house is already energy efficient by Energy label 63
Table 30 Goodness of Fit 64
Table 31 Likelihood Ratio Test 65
01
INTRODUCTION TO THE
RESEARCH
“We are on the eve of the largest transformation. The transformation of our 7 million houses and 1 million buildings, mostly moderately insulated and almost all heated by natural gas, into well-insulated homes and buildings that we heat with sustainable energy and in which we use clean electricity or even generate it ourselves.” (Klimaatakkoord, 2018)
So reads the first sentence of the new Dutch Climate Agreement (2018). Climate change is an important reason for this renovation. But there is more. The Dutch cabinet has decided to stop the extraction of natural gas in Groningen as quickly as possible (Ministry of Economic Affairs and Climate, 2018).
This transformation is a huge task. In order to reach the goal of 3,4Mton less CO² emission by 2030 (compared to the reference scenario) and energy neutral by 2050, roughly 50.000 houses per year require sustainable
improvements and by 2030 almost 200.000 houses per year (Klimaatakkoord, 2018).
The programme “Wijkgerichte aanpak” set up by the Ministry of Economic Affairs and Climate (2018) must make a major contribution towards this goal. The aim of the “Wijkgerichte aanpak” is to formulate a feasible approach for each neighbourhood where the participation of citizens, companies, building owners and other stakeholders is guaranteed. Municipalities play a crucial role in this. Together with all the stakeholders, municipalities have to consider what the best solution is for each neighbourhood, if houses are no longer heated by the traditional heating system. By the end of 2021, municipalities have to present a time-line for sustainability showing which
01.01 The context of the problem
The neighbourhood approach is not new. Programmes such as 'Meer met minder ' and 'Blok voor Blok' have already been set up to increase the number of energy transition initiatives in the existing housing stock, but the findings (RVO, 2014) have shown that it is difficult to entice large numbers of owner-occupied homes to take energy-saving measures. The projects show that owners of owner-occupied homes do not opt for a serial approach, whereby entire streets take the same measures.
Projects in the private sector also appear to be hardly or not profitable at the moment because it takes a lot of "energy" and time to entice owner-occupiers to take energy-saving measures (RVO. 2014). The parties that are active in this are mostly financed by the government or invest in this for other reasons, such as customer loyalty (energy companies) and anticipated lower costs in the future (network operators).
According to K.Raats (2017), the reason for the failure is the object-orientation approach on which these programs are based. In the object-orientation approach, a distinction is made between the subject (the planner) and the object (the space). After analysing the space, solutions are devised by planners (often in the form of a specific technique, such as wind, sun and geothermal heat). This can be qualified as a top-down strategy in which the planners (subjects) arrange the space (object) according to their expertise.
However, the planning system of the Netherlands is based on a sociocratic perspective, in which the rights of citizens are protected and the trade-off between different interests is guaranteed. The object-orientation approach takes these principles only very limited into account. As a results, resistance and frustration grow, delays occur and projects fail.
The "new" neighbourhood-oriented approach is much more based on the sociocratic approach so that it fits the current planning structure and culture of the Netherlands. From this perspective, spatial planning is a trade-off between different meanings given to reality: that of experts, energy providers, citizens and other stakeholders. It is the task of planners to bring these worlds together and achieve successful implementation of new plans.
Participation is a powerful means in this approach. Participation can make a major contribution to the acceptance of the energy transition. However, participation does not automatically lead to acceptance. For both
communication and participation, it is very important to have a good profile of the neighbourhood, both technically and socially / culturally (Klimaatakkoord, 2018). The profile of the neighbourhood can be input for the form of participation to be chosen, such as an information model, participation model, consultation model and
co-production model. The profile of the neighbourhood can also give direction to the communication message and the associated means of communication.
According to Murphy & Cohen (2001), “participation also increases the awareness and engagement of the public in decision-making processes for environmental protection and will ultimately, strengthen that protection” (p.30). Besides, public participation may also result in behavioural change by consumers. At a minimum there is the hope that an engagement of consumers will mean a greater awareness by consumers of the environmental impact of their purchases and behaviour (Barry, 2006).
However, up till now there is little know about the different segments within the energy transition and is limited to a number of pilot projects ‘Proeftuinen Aardgasvrije Wijken’. More knowledge regarding the different profiles of a neighbourhood in relation to energy efficiency behaviour is therefore necessary.
01.02 Research aim and research questions
Drawing on the theory of market segmentation, this research paves the way for a participation model by exploring various segments among Dutch homeowners regarding energy efficiency behaviour. And to further investigate them in order to provide a clear picture of the diversity of Dutch homeowners with regard to energy efficiency behaviour.
In this study, attention will be drawn to the motivation of homeowners towards energy efficiency renovations (EERs). The reason for this is that energy efficiency renovations in the first place are determined by someone’s motivation to do so (Wilson,2015). A segmentation of homeowners based on mutual motivations is most valuable for policy-makers and professionals since motivations can be influenced to a certain extent by communication messages and policy instruments.
Using statistical analysis, segments (a segment refers to a group of homeowners who differ in terms of socio- demographic and house characteristics from others) are determined. These segments are therefore defined from a theoretical and scientific background. A point of attention is that the results also need to be relevant to policy. In other words, this thesis is searching for a set of statistically determined segments that are identifiable, limited and workable, where each of the segments forms a clearly defined and interpreted market segment. This information will help policy-makers and professionals to consider 1) the segments of homeowners most/least likely to invest in energy efficiency renovations 2) the types of programmes likely to be most attractive to the different segments, and 3) the most compelling messages for promoting energy efficiency programs.
In order to gather this knowledge, the research goal is the following:
This research aims to get more insight into different segments in the context of energy efficiency renovating behaviour and of proposing effective communication messages in order to reach the targets of the Climate Agreement .
The aim of this research is translated into the following research question:
► To what extent do personal and contextual factors influence motivation, which is considered to be the antecedent towards energy efficient renovations?
In order to answer the main research question, four sub-questions are derived from it:
► Why do owner-occupiers perform EERs?
► What are the motivations towards EERs in the owner-occupied housing stock?
► What are the (personal and contextual) factors influencing owner-occupiers’ motivations for EER?
► Which segments can be distinguished within the private housing stock and what are effective communication messages for each of the segments?
The aim of this research is not to explain the relationship between personal and contextual factors on the one hand and motivations towards EERs on the other. We are merely interested in the characteristics of households in relation to their motivation towards EER.
01.03 Defining Energy efficiency Renovations
This section gives a description of the concept of Energy Efficiency Renovations (EERS). A better understanding of this concept is necessary to determine in chapter 3 who belongs to the group of renovators and non-renovators.
The term Energy efficiency Renovations (EERs) refers to the range of energy saving measures (energy efficiency) and the way in which these measures are implemented (Renovations). These two together form the definition of EERs. According to Dixion (2013) “the term renovations refer to major structural improvement work to a domestic property, i.e., substantial improvement to a building” (T. Dixon, 2013. P,499) ). Meijer et al. (2009) use renovation to cover renovation actions that go beyond mere maintenance of the building stock. “Renovations have high costs, and skill requirements, and are typically carried out by professional contractors with appropriate technical expertise” (CJ. Maller, 2011. p.2).
The energy-efficient renovation of buildings may address both renewable energies systems (heating or hot water systems) and energy efficiency improvements of the building shell (windows, doors, cavity or loft insulation) (T. Dietz, 2009).
According to the European Commission, there are three types of energy renovations: the implementation of single measures (including the low-hanging fruit), the combination of single measures (which can be termed standard renovation), and the deep or major energy renovation—referring to renovations that capture the full economic energy efficiency potential of improvements (European Commission, 2014).
Besides, energy efficiency renovations can be both undertaken by individuals or groups. If energy efficiency renovations are undertaken by a group of people, the term Local renewable energy initiatives (L.R.E.I) is used. The L.R.E.I. concept is often related to the community concept, and refers to these sorts of initiatives as community energy initiatives (Smith, 2008).
In this research we are merely interested in the installation of renewable energy systems (e.g. installation of a heating system based on renewable energies) and energy efficiency improvement of the building shell. Smaller measures such as replacing traditional lighting for LED and local renewable energy efficiency initiatives are outside the scope of this research.
01.04 Defining the Dutch Home owners
The section gives a description of the definition of homeowners. A better understanding of who belongs to the group homeowners is necessary to determine the research population in chapter 3.
Three categories of private homeowners can be distinguished: Owner residents, Owner associations and Private investors (platform 31, 2016).
Owner Residents
Owners are responsible for the quality of their home. This involves regular maintenance to maintain quality and improvements to adapt quality to new wishes and circumstances. That is most obvious for owners in a detached house. For houses in one building block, agreements will have to be made with the neighbours about certain types of maintenance. An example are apartments in need of a new roof.
Owner associations
Apartments have an owner association (VvE). Decisions regarding energy efficiency renovations are made by the owner association (platform 31, 2016).
Private Rent
In addition to owner residents, part of the private housing stock is owned by private investors, whether or not part of an owner association. Private investors own one or more houses. They rent out properties for profit. They appear to have their own motives for whether or not to invest in home maintenance. Important motives for investing are value development and rental income (cash flows) in the shorter term (platform 31, 2016).
Within the definition of this research, homeowners are considered as owner residents not part of an owner association. Owner residents are entirely responsible themselves for carrying out the renovations of their home whereas the decision of owners of apartments (with an owner association) is made by the owner association (VvE). Private investors seem to have their own motives to invest in home maintenance and therefore excluded in this research as well.
01.05 Scientific and societal relevance of the research 01.05.01 Scientific relevance
Applied research into energy efficient has identified a wide range of explanatory variables describing why homeowners may be motivated to renovate or why these motivations may be thwarted. Broadly these can be divided into drivers and barriers.
The research on energy efficiency renovating behaviour have evolved over time from an emphasis on (financial) barriers and drivers to include a number of variables describing renovation decision makers and decision contexts (Wilson et al., 2014). “These two categories of exogenous influence on the decision correspond to the distinction in social psychology between personal and contextual influences” (Stern, 2000. p.30).
Poortinga et al. (2003) controlled for socioeconomic variables and environmental attitudes in their analysis of UK household preferences for efficient heating systems and insulation measures. Jakob (2007) and Grosche & Vance (2009) tested the influence of household and property characteristics on the adoption of different home efficiency measures in Switzerland and Germany respectively. Braun et al. (2010) also modelled heating system purchase decisions as a function of property and household characteristics, but extended the set of control variables to include location and home tenure. Michelsen and Madlener (2012) include technology attributes as well as home and spatial characteristics in their modelling of renewable heating system choices in Germany.
Although some researchers have examined the factors influencing the decision making process of energy efficiency renovations, this body of work is still limited. By examining how background factors differ between segments within the private housing stock, this paper fills the literature gap and enables policy-makers, professionals and scientists to distinguish groups within the energy transition and formulate more comprehensive policies and information campaigns. Further, most research on energy efficiency is not directed to the case of the Netherlands. It is therefore uncertain if results from domestic research also apply to the case of the Netherlands. This research tries to achieve more clarity on this matter. Last, most research on energy efficiency behaviour is based on intentions of homeowners rather than actual behaviour towards EERs (surveys asking respondents whether they would consider making improvements to their home to make it more energy efficient). Findings based on actual behaviour are more vulnerable since intentions do not always lead to actual behaviour.
01.05.02 Social Relevance
The central aim of the Climate Agreement (EZK, 2018), the reduction of greenhouse gas emissions, touches on our everyday life and that of future generations. Everyone benefits from a safe and healthy environment, sooner rather than later. More knowledge about the renovators and non-renovators as well as the different segments that exist among Dutch homeowners concerning energy efficiency renovations, therefore not only contributes to the scientific field, but also to the achievement of a healthy and liveable planet.
The energy transition can be considered as one of the biggest developments in the coming decades. Decisions made on national and local level impacts on how we live and how we spend our money. For the energy transition to be successful, it is therefore essential that all homeowners feel heard, seen and understood. Communication messages and policy instruments should cover all these aspects. This will increase the number of energy efficient renovations and so the achievement of a liveable environment. Besides, the process will be much better due to mutual understanding and so creating a extensive cooperation between governments and service providers on the
interaction cause economic loss and conflicts. Even the slightest progress in this matter can make our planet, besides a healthier place, also a better place to life.
01.06 Outline
The remainder of this research is structured as follows: Chapter 2 examines the relevant literature on energy efficiency renovations and market segmentation. In chapter 3 the conceptual model will be filled in and the researched methodology will be discussed. Chapter 4 presents the results of the empirical research. Chapter 5 concludes on the analysis and discusses some policy implications.
02
LITERATURE REVIEW
As mentioned in previous section, It is poorly understood which segments exist within the private housing stock in relation to energy efficiency renovating behaviour. We argued that a segmentation of households based on common motivations provides vulnerable information for policymakers and professionals.
Before we can analyse to what extend personal and contextual factors influence motivations towards EER, we first need to understand why owner-occupiers perform or not perform EERs. In other words: What are the factors influencing owner-occupiers’ motivation? This chapter aims to provide some clarity to this through a review of the existing literature and applicable motivation theories.
02.01 Motivation concepts
Academic literature has defined ‘motivation’ in a number of ways. According to Heckhausen and Heckhausen (2008), “motivation is concerned with activities reflecting the drive to attain specific goals” (p.4). Grothmann and Reusswig (2006) define motivation as the intention and willingness to act. This indicates that homeowners first must be motivated to carry out EERs before undertaking an actual refurbishment (Proverbs and Lamond, 2008).
A number of motivation theories have been developed over the past century in order to explain the complex nature of motivation. These have typically “been based on rather divergent and often extreme images and assumptions about the nature of humans” (Ford, 1992, p.8). “Early research on motivation have assumed humans to be
biologically driven ‘robots’, whose actions are dictated by ‘animal instincts’, whereas later research have assumed a logical man, portrayed as ‘scientists’” (S. Organ, 2015. p.39). Recent motivation theories have been based on a combination of such ideas. An exhaustive summary of all relevant theories is beyond the scope of this paper since it is not our aim to explain the relation between personal and contextual characteristics on the one hand, and motivation towards EERs on the other in great detail. However, some basic understanding of this relationship is necessary. In the next section we will briefly discuss the model of S. Organ (2015) which is based on a critical review of the existing literature on motivation theory and its application to EERs in the owner-occupied housing stock. The theories include Vroom’s Expectancy Theory, Maslow’s Hierarchy, Self Discrepancy Theory and Festinger’s Cognitive Dissonance Theory.
According to Organ (2015), “existing motivation theories suggest that people are driven by a combination of needs and desires, their expectations and perceptions of outcomes, their perception of risk, their perception of actual and ideal ‘self’, and by social norms” (p.44). Organ (2015) argues that this is useful in providing some insight into owner-occupier motivations for EER. However, Organ (2015) argues that “existing models are not sufficiently applicable to provide adequate understanding” (p.44). The model provided by Organ therefore incorporates concepts from existing theories, but unlike many theories, the model relates the factors to the context of EERs. Organ (2015)
Figure 1 Model of the internal and external factors affecting owner-occupier motivation for EER (S. Organ, 2015)
Organ(2015) start explaining the Internal factors, which are in the centre of the model. These represent the owner-occupiers’ attitudes, beliefs and values, their locus of control, their sense of responsibility, their perception of ‘self’ and their perception of the role of the ‘home’ etc. “How decisions are made in relation to the home is dependent on how owner-occupiers perceive the role of the ‘home’ (i.e. is it a platform for activities, social interactions, a haven, and so on?), their priorities, their perception of ‘self’ (including actual, ideal and social image), social norms (i.e. what they perceive as acceptable by the social group to which they belong or aspire to belong), and fashions and tastes at the time” (Organ, 2015.p. 86). These and their propriety are specific to each individual, and interrelate and change over time. This is reflected by the arrows circling the internal factors in Figure 4.2. For example, The locus of control of an individual may also be affected by their attitudes, beliefs and values. The internal factors affect the motivation towards EER as well as the form that action takes (decision to renovate or not to renovate). This is indicated in Figure 4.2 by the arrows flowing from the internal factors to EER motivation and action.
Organ (2015) argues that “Internal factors, energy efficiency motivation and EER action will all be affected by external factors such as current incentives, penalties, social norms, the housing market, the condition of the property, socio-demographic characteristics, the cost of the works, regulations, amongst others” (p.87). This is shown in figure 1 by the large arrows moving from the ‘external factors’ layer, inwardly through the other layers.
While we use this integrative conceptualization as our overarching framework for understanding motivation towards EERs, this research only centres on a subset of the model: Socio-demographic and property characteristics (External factors). For the purpose of this research it is sufficient to conclude that external factors influence internal factors, which in turn influence motivation and action. This indirectly implies that motivations may differ between households since the characteristics of households (the context in which households live) is diverse. Which internal factors cause the difference between households motivations is irrelevant for this research since the aim of this research is to identify segments within the private housing stock and not to explain the relation between personal and contextual factors and motivation towards EERs.
02.02 Market segmentation
Before we answer the questions What are the motivations towards EERs in the owner-occupied housing stock and
what are the (personal and contextual) factors influencing owner-occupiers’ motivations for EER?, we will first briefly
discuss the concept of market segmentation.
Smith (1956) refer to market segmentation as a marketing strategy referring to the process of dividing a potential market into distinct subsets of consumers who have common needs and priorities and selecting one or more segments as a target market (p.18). According to Smith (1956) “the basic idea behind market segmentation is that consumers are not all the same and that one strategy will not work for all customers” (p.19).
In this research, the market (Dutch homeowners of privately owned dwellings) will be segmented based on common motivations towards energy efficiency renovations. However, the segmentation itself is based on a limited number of characteristics. Four commonly seen, e.g., Beane and Ennis (1987), Wedel and Kamakura (2000), Doornbos (2004), and Schiffman and Kanuk (2010). Those are listed below.
Table 1 segmentation characteristics
Characteristic Description
Demographic characteristics age, gender, ethnicity, family composition, education level, housing type, and income.
Geographical characteristics City, region, province, postal code.
Psychographic characteristics, or lifestyle characteristics Activities, interests, opinions, attitude, and values.
Behavioural characteristics Motive to buy/benefits sought (price, esthetic, functionality, idiosyncratic preferences),
Considering the motivation theory of Organ (2015), a segmentation based on psychographic or lifestyle characteristics is most vulnerable since attitudes, values and beliefs directly impact households motivations. However, these are not easily identifiable characteristics. That is why it is often examined which motives have a high or low score between groups; this makes groups easier to identify (one of the requirements of section 1.3). We will therefore segment homeowners on the basis of personal (sociodemographic) and contextual (property) characteristics. This information is common to municipalities and freely available for professionals.
Next, an attempt will be made to provide an overall perspective of variables, seeking to direct the approach towards the situation of energy efficiency renovations.
02.03 Likely motivated actors
Personal (sociodemographic) factors
According to Hawkins et al (2007), Individuals’ behaviour is influenced by demographic factors such as education, age, and income. Studies show that level of education influences the acceptance of energy efficiency measures (Held, 1983; Olsen, 1983; Ürge-Vorsatz and Hauff, 2001). Homeowners with lower educational level are more likely to take single measures compared to those with higher educational qualification (Poortinga et al., 2003).
Homeowners’ age influences their energy efficiency behaviour (Carlsson-Kanyama et al., 2005; Mahapatra and Gustavsson, 2008). However, the findings varies across studies. According to Mahapatra and Gustavsson (2008), older homeowners are less likely to adopt energy efficiency investment measures. “this may be linked to their perceived uncertainty of whether the investment will be paid back during their occupancy in the house, less concern about the energy situation, and lesser awareness about energy efficiency measures” (p.105). However, Long (1993) reported that elderly (>65 years) homeowners in America made significant investments in energy efficiency measures. According to Barr et al. (2005), respondents with a mean age of 55 years were more likely to to carry out energy efficiency measures compared to younger age groups. According to the results from the British Social Attitudes survey (DCLG, 2011a), people from the age of 35 to 54 years old and 55 to 64 years old were more likely to carry out energy efficiency improvements to the home (75% and 78% respectively), compared to people form the age of 18 to 34 year olds (67%) and over 65 year olds (58%).
The influence of income on the adoption of energy efficiency measures varies across studies as well. Some studies found a significant association between household income and investment behaviour (Bartiaux et al., 2006; Black et al., 1985; Costanzo et al., 1986; Dillman et al., 1983; Herring et al., 2007), while other studies indicate no or a low correlation between income and investment behaviour (Barr et al., 2005; Ruderman et al., 1987; Ürge-Vorsatz and Hauff, 2001). A recent study identified a strong differences in the adoption of energy efficiency technologies by income groups in eight European countries. Lowest income groups has less willingness to invest for all types of energy efficient technologies (Schleich, 2019). higher income groups are more willing to consider energy works. According to the results from the British Social Attitudes survey (DCLG, 2011a), owner-occupiers with lower incomes (< £12,000) are less likely to consider energy efficiency improvements of the come (58%) compared to households that earn more than £44,000 (86%).
Last, we found some literature on the association between household type and investment behaviour. According to the results from the British Social Attitudes survey (DCLG, 2009b), the group most likely to undertake home improvements are those with growing families and two adult households. The personal factors influencing EERs are summarized in table 2.
Table 2 Personal (sociodemographic) factors influencing EERs
Socio-demographic factors Education Age Income Ethnicity Household type
Contextual (property and neighbourhood) factors
Households living in semi-detached or detached properties are more likely to invest in energy efficiency measures (Caird et al., 2008). Gustavsson and Joelsson (2007) argued that “age of the house could influence the adoption of building envelope measures” (p. 50). Based on their study in Sweden, they argued that owners of older houses may be more inclined to adopt such measures because old houses may be in poor condition, requiring the installation of new building envelope components. According to van Hal et al. (2008) homeowners are more willing to invest in EERs if they have the opportunity to cooperate with neighbours. However, van Hal et al. argues that “the
willingness to contribute to the community depends on citizens’ social connections to their community or a specific institution. Having a strong identification and connection strengthens citizen collaboration and action” (p. 24). Last, Galster and Hesser (1982) elaborate on how the physical condition of the neighbourhood influences the
homeowner’s maintenance and improvement decisions. Galster and Hesser (1982) found empirical evidence showing that the physical condition of the surrounding neighbourhood significantly affect household perceptions of that dwelling’s quality and expectations of tenure and consequently alter maintenance behaviour. The contextual factors are summarized in table 3.
Table 3 Contextual factors (property and neighbourhood) influencing EERs
Physical factors Construction period House type
Dwelling location (density) Neighbourhood cohesion Neighbourhood quality
02.04 Motivations towards EERs
Since motivation is the precursor to EER (i.e. action), in order to encourage the number of energy efficiency renovations, a better understanding of the different motivations is required. This section therefore tries to identify what are the motivations for EERs in the owner-occupied housing stock.
Energy bill savings, to increasing comfort and to reduce environmental impact are three recurrent motivations in the literature on energy efficiency renovating behaviour (DCLG, 2011). These will be discussed in more detail below.
Economic motivation
The principal motivation across all income groups for performing energy efficiency works in the home is to make monetary savings on energy bills (DCLG, 2011; Bichard and Kamierczak, 2009). Besides, increased market value after the renovation is cited in numerous studies to be a important reason for homeowners to carry out EERs (Cirman et al., 2011; Zundel and Stieß, 2011; Organ et al., 2013).
Although individuals tend to prioritise energy bill reduction over motives(Nair et al., 2010), the reason to perform energy efficiency renovations because of monetary savings on energy bills is to a great extend affected by households income. Martinson et al. (2011) argue that “as energy prices continue to rise, it is the lower income owner-occupiers which have stronger incentives to save energy than households with higher incomes” (p.5185).
Social motivation
The second most important reason identified in literature is ‘comfort in the home’ DCLG, 2011a; Pellegrini Masini et al., 2010; Aune, 2007; AnkerNilssen, 2003). Numerous studies show that energy efficiency improvements are performed to increase the comfort of the home, and because of a perceived necessity (Munro and Leather, 1999). The DCLG (2011) report found that comfort of the home as a principal reason increased as income decreased. In addition to the report of the DCLG (2011), comfort has also been found to have greater importance over monetary savings in higher income households (Pellegrini Masini et al., 2010; AnkerNilssen, 2003).
Environmental motivation
The third most import reason identified in literature is ‘environmental concern’. Some studies have suggested that the reason to carry out EERs because of environmental concern is associated with income. Those studies argue that higher income groups are more likely to be environmentally concerned, suggesting that owner-occupiers from higher income groups are more likely to undertake EER for environmental reasons than those from lower income groups. However, Anker Nilssen (2003) concluded in her research that higher income groups tend be less environmentally concerned.
In addition to the motivations described above, owner occupiers also have a number of reasons (motivations) not to carry out EERs. These are cited in numerous studies as barriers to EERs. Below we will discuss the principal barriers to EERs.
High costs associated with an energy efficiency upgrade and not having the necessary financial resources are identified as a major obstacle towards EERs (Black et al., 1985; Jakob, 2007; Nair et al., 2010; Rosenow and Eyre, 2013). The DCLG (2011a) report shows that couples with children were more likely than other household types to feel that the expense of making such improvements is a barrier. Besides, respondents aged 65 and older were less likely than other age groups to indicate the costs of improvements as a barrier.
According to Wilson et al. (2013), conflicting, imperfect or biased information is suggested to be a barrier to EERs. Wilson et al., (2013) conclude in there research that “the Lack of access to competent and credible contractors, previous negative experience, and perceived inconvenience of supervising the contractors are categorized as other important non-economic barriers” (p.33). Further, plans about moving home are found to hinder owners intention to carry out energy efficiency improvements (Matschoss et al., 2012). Disruption to everyday life resulting from the renovation project is found to be a principal barrier towards EERs by Weiss et al. (2012). An barrier that came up in the research of Zundel and Stieß (2011) is that households indicated that up to now, they did not take the time to deal with this decision. Table 4 summarises the motives found in literature.
Table 4 Motives towards EERs
Motives to carry our EERs Motives not to carry our EERs
Motivation theme Sub-theme Motivation theme Sub-theme
Economic ► Potential savings
► Availability of capital
► Value added to property
Economic ► Costs of upgrade
► Insufficient potential savings
► Availability of capital
Social ► Comfort (Health and
safety)
Information ► conflicting, imperfect or biased information
► previous negative experience Environmental ► Carbon footprint Operational ► Disruption
► Social costs (time effort)
► plans about moving home
02.05 Conclusion on the literature
It can be concluded that homeowners motivation is multifaceted and complex. Existing motivation theories suggest that people are driven by a combination of needs and desires, their expectations and perceptions of outcomes, their perception of risk, their perception of actual and ideal ‘self’, and by social norms. Organ (2015) argues that these models are not sufficiently applicable to provide adequate understanding of owner-occupiers’ motivations. Organ (2015) introduced a new model showing how internal and external factors influence homeowners’ motivations and action.
In order to specify and operationalize the motives and personal and contextual factors, we discussed previous studies related to energy efficiency renovating behaviour. We found numerous factors influencing refurbishment ‘decision-making’, and therefore potentially has a role in the motivation for EER.
03
METHODOLOGY
This chapter outlines the research methodology used to investigate the research question. First, the underlying philosophical assumptions are discussed. Second, the research design - including research methods, data collection and analysis techniques - is described.
03.01 Philosophical viewpoint
To best explain the research methodology chosen to answer the research question, a short discussion is necessary. Creswell (2003) describes a methodology as a “strategy or plan of action that links methods to outcomes”, affecting the choice of methods and their use (p.5). Research methodologies are guided by the philosophical assumptions of the researcher. These assumptions relate to the nature of reality (the ontology) and to the extent to which this reality can be known (the epistemology). Philosophical assumptions, epistemologies, ontologies and methodologies have been grouped into research paradigms (Creswell, 2007; Creswell, 2003).
As quite clearly exposed by Thomas (2013), there are two categories of researchers with two different ways of creating knowledge: Martians and the naturalism approach, and Venusians and the constructivism approach. According to naturalism, there is a real world out there that is independent of our senses. Moreover, naturalists assume that the real world exists whether human beings are there to observe it or not, and the truth is identified only when what it corresponds to reality (Moses & Knutsen, 2007). Still, the question “what is reality” arises. In addition, constructivists strongly believe that reality is socially constructed, and that each of us has the capacity to construct his/her own meaning about the world through his/her own senses. As a matter of fact, constructivism assumes that the way individuals experience and interpret reality is directly proportional to their social dynamics, (past) experience and information. Individuals’ constructions are not more or less “true”, in any absolute sense, but simply more or less informed and/or sophisticated (Guba & Lincoln, 1994). Therefore, the constructivist approach encourages the researcher to involve him/herself in the research context, talk to people in depth and attend to every nuance of their behaviour in order to understand them (Thomas, 2013).
Considering that the purpose of this study is to produce conclusion on the existence of sociodemographic groups which differ in terms of motivations towards energy efficiency renovations, it can be concluded that this study design suits best a naturalist position. In addition, a subjectivist (who believes that there are multiple truths) probably want to understand why households may have different motivations towards EERs.
03.02 Research strategy.
In this section, the research strategy for answering the research question To what extent do personal and
There are two types of basic research approaches: qualitative and quantitative. Qualitative research is by definition exploratory. It is used to go deeper into issues and explore nuances related to the problem. Quantitative research is conclusive in its purpose, as it tries to quantify a problem and understand how prevalent it is by looking for
projectable results to a larger population (J.W. Creswell, 2013). Since the aim of this research is to identify groups within the energy transition that have different motivations towards energy efficiency renovations, the quantitative approach appears to be the best (J.W. Creswell, 2013). It is not our goal to find an explanation for the relation between motivation and homeowners’ backgrounds which is more suitable for a qualitative approach.
Quantitative research has a number of important characteristics compared to qualitative research. Quantitative research gives the researcher the opportunity to draw conclusions which can generalized to an entire target group. In addition, quantitative research offers the possibility to analyse similarities and differences between subgroups within the overall target group. However, quantitative research has also limitations. Quantitative experiments do not allow participants to explain their choices or the meaning of the questions may have for those participants (Carr, 1994). Besides, The researcher might miss observing phenomena because of focus on theory. It is important to take these limitations into account when conducting a quantitative research.
There are three ways to collect data for a quantitative research study: Surveys, experiments and observations (J.W. Creswell, 2013). This study uses the data collected during the Housing research 2018 (Woononderzoek). The housing research is a national survey that is conducted every three years. The survey consists of different modulus with questions that are answered by the same person. The surveys for the Dutch housing research were conducted from September 2017 to May 2018. In addition to the national survey commissioned by the Ministry of the Interior and Kingdom Relations, Directorate of Governance, Spatial Planning and Housing, local parties such as
municipalities , provinces and regions had the opportunity to increase the sample for their working area. The fieldwork for oversampling runs parallel to the fieldwork of the national research. The new WoON database contains more than 67,000 successful surveys, of which more than 24,000 on behalf of the oversampling participants. The results were announced on 4 April 2019. This study will be one of the first using the data of the new housing research.
The Dutch housing research offers a number of advantages in comparison to carry out surveys yourself
► The housing survey is unique in terms of size. It is impossible to conduct such a number of surveys over a period of 5 months. The size of the WoON is such that it provides support for reliable judgments at national, provincial and regional level (WoON, 2018)
► When setting up a survey, it is uncertain in advance how many respondents there will be. As a result, there is a risk that the desired number of respondents will not be reached or that the number of respondents will be unevenly distributed among different groups, making it impossible to generalize outcomes.
► Compiling surveys is complicated and requires expertise. Formulating a wrong question means that something else is being measured than what is intended (validity).
There are also some disadvantages related to the use of secondary data (Saunders et al., 2009).
► The questions related to energy efficiency renovations are limited (availability of data)
The advantage of conducting surveys yourself does not outweigh the use of the results from the 2018 housing survey in this case since this research is merely descriptive and aims to identify characteristics and categories of homeowners towards energy efficiency renovations and generalise the results to the entire Dutch private housing stock.
Certain choices within this study will therefore be determined by the availability of data. For example: The housing survey does not contain many questions on attitude towards the environment. Therefore it is not possible to segment groups based on their mutual attitude or beliefs. However, it is possible to divide people in groups based on their motivation to EERs since the questionnaire consists of a large set of questions asking what motivates or hinders homeowners in their decision to EERs.
03.03 Conceptual model
This conceptual model is based on the motivation theory of Organ (2015). The basic model is based on the idea that the behaviour of people is driven by certain motives. Within our research drivers or barriers towards EERs. Motives in turn are indirectly influenced by personal and contextual factors. Motives create behaviour (action). Within this research: The decision of homeowners to carry out EERs or the decision of homeowners not to carry out EERs. The conceptual model is shown in Figure 2.
Figure 2 Conceptual model based on the model of Organ (2015)
In order to fill in the basic model and specify and operationalize the motives, personal and contextual factors, a literature study is performed. The operationalising of the variables will be discussed in section 3.5.
Motives (Drivers)
Motives (Barriers)
Personal and Contextual facors
Renovators NNon-
28 Renovators and non-renovators will be analysed separate in this research due to the limitations of the Housing survey. In the questionnaire, respondents were asked if they have conducted energy efficient renovations in the past five years. If they had invested in EERs, respondents were asked what drives their decision. Respondents who responded negative to the first question were asked what hindered them.
After analysing which personal and contextual factors are significantly different between renovators and non-renovators, statistical analyses will be derived to examine whether or not groups with common motives differ in terms of household, property, neighbourhood characteristics and type of project.
Note: The first part of the analyses examines which factors are significantly related to EERs. In the second part of the analyses, a segmentation of the resident owners will be made based on their motivation towards EERs (indicated by the yellow rectangle). Differences in terms of household, property, neighbourhood characteristics and type of project will be examined.
Figure 3 Conceptual model (2.0)
Empirical indicators
Market segments (motivations) Since the Since the multinomial regression analyses only determines the odds of a category compared to the reference category, we will conduct a Chi-square test for
the significant indicators.multin omial regression analyses only determines the odds of a category compared to the reference category, we will conduct a Chi-square test for
the significant indicators. Property
characteristics Type of project
Household characteristics Neighbourhood characteristics Theoretical Indicators Drivers Since the Since the multino mial regressi on analyses only determi nes the odds of a category compare d to the referenc e category , we will conduct a Chi-Barriers Since the Since the multino mial regressi on analyse s only determi nes the odds of a categor y compar ed to the referen ce categor y, we
03.04 Hypothesis
We found 10 personal and contextual factors related to EERs (table X) in literature. These are reflected in the housing survey as well. The personal and contextual factors found in literate as well as reflected in the housing survey are translated into research hypothesis. This will be tested in chapter 4.
Personal influences: Personal influences concern sociodemographic characteristics. Earlier research shows that these aspects are important determiners of energy efficiency renovation behaviour (section 2.4.3). The hypothesis related to this part of the model are:
Hypothesis 1 Age is associated with EERs (older households are more likely to invest in EERs) Hypothesis 2 Ethnicity is associated with EERs (Natives are more likely to perform EERs)
Hypothesis 3 Education is associated with EERs (higher educated are more likely to perform EERs than less well educated)
Hypothesis 4 Income is associated with EERs (higher income households are more likely to invest in EERs) Hypothesis 5 Household type is associated with EERs (two-adult households are more likely to perform EERs)
Contextual influences: This concept is split into seven hypotheses.
This is because earlier research shows that these aspects influence the decision of homeowners towards energy efficiency renovations (section 2.4.1). The hypotheses related to this part of the model are:
Hypothesis 6 Building age is associated with EERs (households living in an older houses are more likely to invest in EERs)
Hypothesis 7 House type is associated with EERs (households living in semi-detached and detached houses are more likely to invest in EERs)
Hypothesis 8 Perceived quality of the neighbourhood is associated with EERs (households perceiving the quality of the neighbourhood as positive are more likely to invest in EERs)
Hypothesis 9 Social cohesion is associated with EERs (households with close bonds to their neighbours are more likely to invest in EERs)
Hypothesis 10 (location is associated with EERs (households living in suburban settings are more likely to invest in EERs)
We found 3 principal reasons in literature why homeowners do carry out EERs. Besides, we found a number of reasons why homeowners do not carry out EERs. Although there is less evidence on the relation between motives and personal and contextual factors, some researchers did analyse this relation. The results can be translated into the following research hypotheses:
Hypothesis 11 Income is associated with motivation (higher income households are less likely to be financial motivated)
Hypothesis 12 Age is associated with motivation (Older households are less likely to be environmental motivated) Hypothesis 13 Household type is associated with motivation (Couples with children are more likely than other household types to feel that the expense of making such improvements is a barrier)
03.05 Research method
In pervious section the research hypothesis are formulated based on the literature study of section 2. This section describes the data used to test the hypothesis, the operationalization of the factors and the selection of the analyses technique.
03.05.01 Data
As mentioned before, the data used in this research is obtained from the Dutch Housing survey (WoON 2018). This survey, carried out by the Ministry of the Interior and Kingdom Relations (BZK) and Statistics Netherlands (CBS), provides every three year insights into the Dutch housing market.
The target population of the WoON survey consists of persons aged 18 years or older, living in the Netherlands on 1 January 2018 and part of a private household. More than 110.000 people have been contacted to participate in WoON 2018. The Housing market module consists of a regular part and an oversampling part. Based on the regular part of the Housing Market Module, reliable statements can be made at national, provincial and COROP level. To increase the usability of the WoON, local governments, provinces, municipalities, regions and housing associations had the possibility to participate in the research via oversampling. This research covers both the regular part as well as the oversampling of the Housing Market Module.
A number of preconditions have been set by Statistics Netherlands together with the Ministry of the Interior and Kingdom Relations. Those include agreements on the level of response, the sample design, the approach strategy and various quality requirements. The preconditions set by Statistics Netherlands together with the Ministry of the Interior and Kingdom Relations have all been met (WoON, 2018).
Table 5 summarizes the status of all persons from the sample when the fieldwork was completed. A total of 44,480 questionnaires were fully completed via the internet (CAWI). That is 65.9% of the total number of responses realized. 13,662 and 9,381 complete responses were achieved via telephone (CATI) and face-to-face observation (CAPI), which corresponds to 20.2% and 13.9% of the total response.
Table 5 Response WoON2018
03.05.02 Non-response analysis
Non-response is a major problem with sample research. During data collection, the aim is to achieve the highest possible response and the best possible representation of the target population. Due to selective non-response, certain groups are under- or overrepresented, which means that the respondents do not accurately reflect the population. This can lead to bias in the results of the study. Weighting is a technique to reduce bias due to non-response. By assigning weights to the respondents, representativeness of the database is restored as good as possible by Statistics Netherlands.
Before composing the database of all the responses, various checks and consistency checks were carried out. These checks can be split into three parts:
► Completeness checks;
► Checks for inconsistencies within the questionnaire;
► Checks for inconsistencies with regard to register data.
After checking the completeness and consistency, the data is enriched with data from, among others, the Personal Records Database (Basisregistratie Personen) and the Tax Authorities (Belastingdienst).The result is a file with 67,523 cases that are assessed as qualitatively good and acceptable for further analysis (Onderzoeksdocumentatie WoON2018). The results are processed in a SPSS data file. This is also the data file that serves as the basis for this secondary analysis. A more detailed explanation of the method of data collection can be found in the report (Onderzoeksdocumentatie WoON2018).
03.05.03 Operationalising the variables
In this section we will discuss how each of the factors will be measured. For objective variables, we will describe the response categories from the questionnaire. Subject variables will be transformed into measurable variables by comprising variable questions from the questionnaire. The score on the different variables will then be summed up. Since the analysis is divided into two sections, first the dependent and independent variables of the first part will be discussed. Second, the dependent variable(s) of the second part of the analysis will be discussed. The independent variables are the same.
Dependent variable
Energy efficiency renovations (EERs): We define the renovators and non-renovators using the following
question 11.1 of the database: Have you installed in the past five years 1) double glazing 2) improved the insulation of walls or floors 3) Installed or replaced solar panels 4) replaced the heat system 5) installed other energy
efficiency measures 6) non energy efficiency renovations. Respondents who installed one or more energy efficiency measures are renovators in this research. Non-renovators did not install any energy efficiency measures in the past five years (category 6).
Independent variables
Personal influences: With regard to the personal influences, this research examines: Age, Education,
Ethnicity, Income, and Household type. The database divides respondents into 7 categories based on their age. The first category contains respondents between 17 and 24. The following categories have a range of 10 years. The last category contains respondents of 75 and older. The question about education divides households into three categories: less well educated, middle educated and highly educated. Income is added to the database by the Dutch statistics (CBS). They divide households into five categories based on their income: Less than average, 1,5 times average, 2 times average, 3 times average and more than 3 times average. The question about ethnicity divides homeowners into three categories: Natives, non-western and western. The question about Household composition divide households into 7 categories: 1)Partners, 2)partners with children, 3)partners with children and others, 4)partners with others, 5)one person household with children, 6)one person household with children and others, 7)different composition.
Contextual influences: With regard to the contextual influences, this research examines: Building age,
1) apartments, 2) Terrace house 3) semi- detached 4) detached 5) home with workspace 6) other housing type. Perceived quality of the neighbourhood is measured by the statement: The houses in this neighbourhood are well maintained. Respondents were asked whether they agree or disagree with the statements based on a 5-point Likert scale. Social cohesion is added to the databased by Statistics Netherlands. The score varies between 0 and 10 whereas 0 means no cohesion and 10 very much cohesion. Location is measured based on the number of houses per square kilometre. The categories are: More than 2500 houses per square kilometre, 1500 – 2500 houses per square kilometre, 1000 – 1500 houses per square kilometre, 500 – 1000 houses per square kilometre and 500 or less houses per square kilometre.
For the second part of the analysis, motivations have been defined based on the questions of the housing survey asking respondents why they performed EERs or why they did not carried out EERs
Motives (Drivers): Respondents who did invest in energy efficiency renovations in the past five years were
asked what drives them. Six answers are possible whereby the respondent can select more than one answer. The categories are: 1) This was necessary because of maintenance. 2) The investment will be earned back due to lower electricity costs. 3) Increase thermal comfort. 4) Environmental concern. 5) This is decided by the housing
association. 6) Make the house easier to sell. Category 5 has no responses since we are only interest in home owners not being part of a housing association. Table 6 shows the category of response corresponding to the driver found in literature.
Table 6 Response category linked to driver
Motives (Barriers): Respondents who did not invest in any energy efficiency renovations in the past five years, were asked what hinders them. Seven answers are possible whereby the respondent can select more than one option. The categories are: 1) The house is already energy efficient. 2) I cannot effort a renovation. 3) The return on investment is considerably low. 4) I do not know what the options are. 5) I do not want a renovation. 6) No time/ I did not think of it. 7) I want to move. Table 7 shows the category of response corresponding to the barrier found in literature.
Table 7 Respond category linked to barrier
DRIVERS Respons catergory
Cost savings 2
Environmetal Benefits 4
Thermal comfort 3
Make the house easier to sell 6
Life events (maintanance) 1
BARRIERS Respons catergory
High capital costs 2
Aversion to delayed gains (high implicit discount rates) 3
Imperfect or biased knowledge of energy costs 3
Invisibility of energy use and/or efficiency measures (e.g., cavity wall insulation) 4
Low % cost of household budget 2
High transaction cost of information search 7
Complexity of decision (information processing) 4
Anticipated disruption to domestic life from renovation work 5
In the questionnaire, respondents can select more than one reason not to carry out EERs. Because, it is unclear which is the most import barrier, it is decided to analyse only the respondents who did select one of the barriers (see section 4.3.1).
03.06 Method selection
This section discusses the various statistical tests that will be used to test the hypotheses and to analyse whether segments differ in terms of personal and contextual indicators.
We explored the use of predefined segmentation models such as VALS, MindBase, or the like. However, these models emphasis on values, attitudes and beliefs. Therefore these models are mainly used to determine segments in terms of lifestyles. Geo-demographic segmentation involves a combination of geographic and demographic data. However it automatically assumes that people who live close to one another are likely to have similar financial means, tastes, preferences, lifestyles, and consumption habits (Schiffman and Kanuk 2010). Statistical analysis techniques in addition enables the researcher to select variables based on the aim of the research and mostly used in housing market research (Visser and van Dam, 2006). This method is therefore preferred in this research. In the next section the statistical tests will be discussed in more detail.
03.06.01 Multivariate analyses
To determine whether renovators and non-renovators as well as segments within these groups differ in terms of personal and contextual characteristics, a Chi-square test will be performed. This method is used to test whether groups differ from each other (when both the dependent and independent variables are nominal). The Chi-square shows if the association between the dependent and independent variables is significant. However, the Chi-square test do have some disadvantage. It does not show how strong the effect is and is does not take the effects of other variables into account (Laerd Statistics, 2015).
Therefore a multivariate analyses will be performed to correct for independent variables that have overlap, for example age and income. It is possible that age and income are significant if they are tested separately (for example with a Chi-square test), but if they are tested together, either age or income is no longer significant because part of the income is determined by ones age.
For the first part of the analysis, we will use a logistic regression technique. In addition to linear regression, with a logistic regression the dependent variable in this analysis consist only of two categories. In this research, the dependent variable is energy efficiency renovations with answer categories ‘yes’ and ‘no’. In this regression, the probability of an event, occurring for randomly selected observations are determined by any given combination of independent variables (Cohen et al., 2014). Table 8 is an example of a logistic regression in Statistical Package for the Social Sciences (SPSS version 25.0).
