1. Introduction
The built environment is a critical sector in terms of its influence on the economy, society, and natural environment as construction activities are estimated to form about 9% of the European gross domestic product [
1] and are the major consumer of natural resources [
2]. Research suggests that this industry is responsible for 39% of global energy-related emissions [
3] and 46% of the total waste generation in the European Union (EU) [
4]. Thus, there is an urgent need for transforming the built environment to a resource-effective one to address these challenges.
The concept of Circular Economy (CE) has been embraced as an approach for minimising resource inputs and outputs by introducing cyclic principles [
5], avoiding waste and pollution, and creating regenerative systems [
6]. The concept gained traction in Europe in the early 2010s with the efforts of Ellen MacArthur Foundation (EMF) along with the introduction of the first Circular Economy Action Plan [
7,
8]. Indeed, many European countries [
9], including the Netherlands [
10], have developed several strategies and action plans, in which the construction sector takes a pivotal role as one of the main priorities in the transition towards a CE.
Research on CE in the built environment covers various dimensions, with some researchers focussing on the material innovation while others address CE implementation at city scale. For example, Marie and Quiasrawi [
11] studied the properties of recycled aggregates that are reintroduced in the concrete life cycle multiple times, van Stijn and Gruis [
12] proposed a circular housing retrofit strategy for modular building components, Eberhardt and colleagues [
7] conducted a systematic literature review to determine which building design and construction strategies are associated with circularity for new buildings, and Prendeville and colleagues [
13] investigated how six European cities are implementing CE as a strategy. Furthermore, several researchers have proposed tools [
14,
15,
16] and assessment methods [
17] to support circular building processes, while others conducted systematic literature reviews to demonstrate the state-of-the-art of CE research [
18,
19] and identified barriers [
20] for CE implementation in the built environment.
However, only a very few of the reviewed studies explicitly examines the circular transition of the housing sector, with a notable example [
21]. This can be considered as somewhat surprising, given that the housing stock constitutes a significant part of the built environment. Moreover, especially in Northwestern Europe, a large part of the housing stock, varying from 3% to 30% of the total housing stock [
22], is managed by professional institutes, social housing organisations (SHOs), with substantial portfolios that offer opportunities to generate circular flows of materials at the portfolio level. For a wider adoption of the CE in the built environment, therefore, understanding of SHOs’ experiences with the circular practices is critical.
Sustainability of social housing is one of the five top priorities of Aedes, the umbrella organisation of Dutch housing associations [
23]. Dutch SHOs own 29% of the national housing stock [
24] and provide services to approximately 4 million low-income residents [
25] which make them prominent actors in the Dutch construction sector.
Based on this background, this article aims to identify (1) circular practices of the early adopter Dutch SHOs; (2) main barriers that hinder CE implementation; and (3) potential enablers to address the most pressing barriers by conducting a Delphi study with 21 sector professionals across the Netherlands.
The remainder of this article is organised as follows.
Section 2 presents the background of the study, discussing relevant literature on CE in the built environment, the main characteristics of Dutch SHOs, and CE implementation barriers and enablers in the construction sector.
Section 3 demonstrates the execution of the Delphi method and elaborates on the data collection and data analysis phases. Further,
Section 4 presents the research results highlighting priority issues, while
Section 5 includes the discussion and concluding remarks.
3. Delphi Method
Delphi is a method for aggregating opinions from a group of knowledgeable individuals for a wide variety of purposes including issue identification, concept development, group decision making, and forecasting future trends [
64,
65,
66]. Early applications of Delphi concern forecasting in the military context; later, it became a popular method, both in academia and the corporate world, for reaching consensus, decision-making, or policy-making [
67,
68]. This technique is considered convenient for several scientific domains as many scholars applied it in social sciences [
67,
69,
70], housing studies [
71,
72,
73,
74], and also in CE-related inquiries [
54,
75,
76,
77,
78,
79]. Furthermore, some researchers used Delphi technique, similar to this study, to determine barriers and enablers for implementing successful CE-based food supply chains [
79] and for the application of sustainable purchasing and supply management [
80].
The Delphi method has four key characteristics that made it suitable as the core method of this study. Based on the literature [
64,
65,
66,
67,
68,
81], these features can be summarised as follows: (1)
Anonymity: During the execution, participants do not confer with each other as the facilitator controls the process. The aim is to reduce the impact of dominant individuals in group decision making. Additionally, anonymity allows respondents to express their opinions freely without feeling group pressure. (2)
Iteration: The questioning of the participants occurs in several rounds of written questionnaires or interviews so that the panellists can adjust their opinions based on the feedback they get from the facilitator. Throughout the process, participants are actively involved in the debate and influence the questions and outcome. (3)
Controlled feedback: The facilitator regularly transfers information between panellists. After each Delphi round, facilitator delivers feedback in a summary of the statistical values of the group judgements. (4)
Statistical group response: At the final stage of the process, participant responses are formulated statistically and presented numerically, graphically, or sometimes qualitatively to indicate the degree of consensus or disagreement.
We performed a two-round Delphi study between December 2019 and October 2020, comprising three overarching phases, as shown in
Figure 2. The preparation phase concerned the panel recruitment and the preparation of a list of barriers and enablers. The execution phase dealt with the data collection through interviews and questionnaires, and the final phase dealt with the analysis of the collected data.
3.1. Preparation
3.1.1. Panel Formulation
Scholars stress two crucial aspects of the panel formulation in Delphi surveys: expertise of the panellists and the size of the panel. The former is related to the selection of experts who have sufficient knowledge and experience in a specific domain [
82], whereas the latter concerns the ideal number of participants in a Delphi panel. Sossa and colleagues [
83] observed a tendency towards using a fewer number of panellists in academic research. Although there is no unique rule for the panel size, it is suggested to keep the participant number between 5 and 20 [
82].
At the beginning of the study, we sent invitations to 64 sector professionals across the country who work for the forerunner SHOs that have explicit ambitions to implement circular principles, and preferably have conducted pilot projects in which they have experimented with circular construction approaches. The selection of forerunner SHOs was made based on reviewing professional journals and sector-related websites, our own knowledge, and the snowball technique. In return, 26 of the invitees responded to our call positively, a response rate of 40%. Following a round of introductory conversations, a panel was formed with 21 professionals, representing 19 different housing association owning approximately 21% of the social housing stock in the Netherlands. The size and locations of the participating SHOs are shown in
Table 1 and
Figure 3, respectively, and the overview of the panel members is presented in
Table 2.
3.1.2. Extensive List of Barriers and Enablers
Prior to the first Delphi round, we prepared an initial set of CE implementation barriers and enablers, based on the relevant literature [
53,
54,
58,
61,
84,
85,
86,
87], to stimulate the discussions with the panel members during the interviews. Similar issues identified by different scholars were merged and sometimes adapted to the context of this study. For example, we combined “Limited awareness across the supply chain” [
61], “Lack of interest, knowledge/skills, and engagement throughout the value chain” [
84], and “Lack of awareness, understanding, knowledge, and experience with environmental issues” [
87] into “Lack of awareness, knowledge, and experience with the CE.” A total of 56 issues were grouped under six categories, namely,
social and cultural, organisational, financial, sectoral, technical and technological, and
regulatory.
3.2. Data Collection
3.2.1. Delphi Round I
The purpose of the first Delphi round was to explore the CE implementation issues that early adopter housing associations experience with their pilot projects. Before the online interviews, panellists were sent a list of barriers and enablers in a questionnaire format and asked to score each of the matters by importance on a 5-point Likert scale, 1 being “not important at all” to 5 being “extremely important”.
As outlined in
Table 2, 19 panel members out of 21 responded to the online questionnaire and participated in the online interviews. At the beginning of the interviews, panellists were asked open questions regarding circular practices in their organisations. Following this, barriers and enablers in each category were refocussed, and panellists’ initial ratings were discussed in-depth. In the meanwhile, panellists reflected on their responses and supplemented additional points that were not covered in the list. These points were then mentioned in the subsequent interviews to validate whether they were relevant to be brought to the second round. Further, panellists were given a chance to adjust their answers upon discussions before the interviews ended. Upon completion of the first round, a summary of the first cut results, demonstrating the mean scores, the highest, and the lowest ratings, and additional notes of the panellists were reported to all participants.
3.2.2. Delphi Round II
There were two underlying objectives of the second Delphi round: (1) to determine circular principles used in business-as-usual practices and circular pilot projects and (2) to prioritise barriers and identify enabling factors. For the former, we used the R framework proposed by Potting and colleagues [
35] and asked panel members to indicate which of the R principles apply for both their regular activities and circular pilot projects. For the latter, panel members ranked 13 barriers, chosen from the previous round, in line with the priority given by their organisations. The selection of these barriers was made according to the top-rated two scores per category, including an additional issue raised by the panel members (“The building code, rules and regulations hinder reusing building materials”). The reader must note that some of the barriers from the first round were combined to keep the list concise. For instance, “High purchasing costs of new circular materials” and “High purchasing costs of recycled materials” were combined into “High purchasing costs of circular materials (new and recycled).” Finally, participants were requested to propose enablers to address the top 5 barriers they ranked. With this, we aimed to build meaningful correlations between the most pressing five barriers and potential enablers.
3.3. Data Analysis
For the first cut summary, a quantitative analysis was performed to summarise the panel ratings by calculating minimum, maximum, mean scores, and standard deviation values. Standard deviation was used to demonstrate the distribution of responses, in other words, the degree of consensus. A lower standard deviation value indicates a higher consensus. We did not seek a consensus among panel members, but focused on exploring CE implementation issues. Therefore, a consensus criterion was not defined when analysing the results. Similarly, for analysing the second-round results, mean and median scores of the rankings were used to measure central tendency, and standard deviation and interquartile range were calculated for quantifying the amount of variation in rankings. After finalising the data analysis, a summary of the results was reported to all panellists.
5. Discussion and Conclusions
Despite the emerging body of literature in CE in the built environment, existing research has mostly overlooked the housing stock, especially the one managed or owned by the social housing organisations (SHOs), while this offers tremendous opportunities to generate circular flows of resources in the built environment. This article sheds light on the CE practices of the early-adopter Dutch SHOs and presents the main barriers and enabling factors associated with implementing circular principles, employing a Delphi study with 21 sector professionals.
Seen from a wider implementation of CE approaches in their maintenance, renovation and construction activities, our findings indicate that Dutch SHOs are at the early stage of development in which they experiment with new circular strategies by involving sector stakeholders from the beginning of the construction process. In doing so, we found a tendency to apply higher level circular strategies, such as “refuse”, “rethink” and “reduce” in pilot projects.
From the circular business models perspective, Dutch SHOs are “service providers” who keep the ownership of the housing stock they operate and offer rental properties to their tenants. This system coincides with “Access and performance model” of Bocken et al. [
5], which was interpreted differently by Eikelenboom et al. [
21] as delivering all-inclusive service package to the tenants through a single contract. They argue that such model could cause extra burden on low-income households. SHOs also regularly repair and maintain their housing stock, slowing the resource loops by offering long-lived buildings, as in “Classic long-life model” [
5]. Therefore, elements of a CE are already implicit in their business operations. However, there is a noticeable gap in new business model creation in circular pilot projects. Among 19 represented SHOs, only two of them employ the take-back system, and one of them tests a materials-as-a-service model with a supplier.
Our Delphi research has identified five critical barriers for a wider implementation of CE in the Dutch SHOs, namely, (1) higher priority in other issues; (2) operating in a linear system; (3) lack of awareness, knowledge, and experience with the CE; (4) high purchasing costs of circular materials (new and recycled); and (5) unclear business case.
In general, the main barriers that Dutch SHOs encounter are closely related to their organisational structure and company culture. This finding coincides with the Kirchherr and colleagues’ EU-wide study [
53]. According to their results, other businesses also suffer from “Hesitant company culture” when introducing CE as a strategic goal in their organisations. On the other hand, Adams and colleagues [
61] discuss organisational issues mainly from the sectoral perspective. Their study with the UK construction industry indicates that the sector’s fragmented nature hinders the application of circular principles throughout the supply chain. The panellists also acknowledged this view in the first round of our Delphi survey. However, we have not observed a direct relationship between the sectoral and organisational barriers.
Similar to our study, several studies highlight that developing a viable business case for circular construction processes is challenging [
61,
94] and high costs of circular materials hamper the CE implementation [
62,
95]. Challenges for new business model creation have ties with the traditional ownership models in the building sector. Several scholars discuss the need for a shift in the way of ownership of buildings and its components are structured for the circular flows of resources [
30,
61,
63,
96]. As discussed previously, Dutch SHOs retain the ownership of their building stock and deliver services to their tenants, which correspond to circular models. However, for renovation and newly built projects, there is a room for experimentation with other circular business models to increase the level of circularity.
Many reviewed studies identify lack of awareness as one of the most critical barriers for CE implementation [
20,
53,
61,
62]. Consistent with the literature, our study also found this barrier very important; however, there is a marked difference in our findings that panel members consider lack of “tenant” interest and awareness as a minor issue, whereas other studies, e.g., Kirchherr and colleagues [
53] found “Lacking consumer interest and awareness” as the most pressing barrier in the European context (see
Section 4.2.1 for panel arguments on this topic).
Several enablers are proposed to overcome these key obstacles. These include a binding CE legislation allowing innovation in circular construction practices and reforming existing tax schemes on construction materials, systematic exchange of best practices, development of enabling technologies and circularity measurement tools, a more prominent role for leadership and priority setting at top-management level, and clear business models for SHOs and their supply chain partners. Particularly for new starters, development of CE design and implementation guidelines and collaborating with other SHOs are important enabling factors.
Overall, our study shows that, although the Dutch SHOs may have been dealt a good hand in terms of their fundamental business model and societal objectives, they also face significant barriers for a wider implementation of CE principles. The main challenge now seems to be setting in place the enablers that will allow circular asset and construction to become common practice.
When interpreting our findings, it must be kept in mind that the Delphi panel members were chosen from SHOs that have explicit goals for the CE. Other SHOs, who have no explicit CE goals yet, may be expected to face similar barriers and enablers when they do start to adopt CE goals, but this cannot be stated with absolute certainty. Moreover, as CE in the construction sector itself evolves over time, the experienced barriers and enablers are likely to shift as well.
This article contributes to the rapidly expanding field of circular built environment research by providing insights from the SHOs, who own a large part of the housing stock, particularly in Northwestern Europe. Our work appears to be one of the first attempts to examine housing associations’ CE practices thoroughly and lays the groundwork for future research into CE implementation in the sector. This study’s findings will be used in further research on the development of a framework to address identified barriers through enabling digital technologies.