Stakeholder Perspectives on Policy, Social, and Organizational Challenges of Sustainable Residential, Multi-Storey Building Retrofitting in Germany

Abstract

Retrofitting existing buildings is regarded as a main driver of decarbonization, yet retrofitting activities are lagging behind their ambitious goals. This study explores 86 German construction practitioners’ perceptions of organizational, policy, and social challenges to sustainable retrofitting and how those perceptions relate to age, attitude, and their interaction. The primary analyses used OLS moderation models with HC3-robust standard errors and ordered-logit models, which served as robustness checks. Across outcomes, more pro-environment attitudes were associated with fewer perceived challenges, and older practitioners (41–56+) reported higher barrier perception. The attitude × age interaction results indicate that the protective link of attitude was weaker among older respondents, which was significant for policy and social challenges but only marginal for organizational challenges. The model fit was reasonable, at an Adj. R² between ≈0.56 and 0.72 with acceptable diagnostics. Our results suggest that even motivated professionals can feel constrained, especially among older, senior staff. Practical implications include early tenant engagement to enhance acceptance and foster internal organizational sustainability capacities. Policy instruments such as adult education programmes need to be leveraged to enhance sustainable construction capabilities and reinforce attitudes and behaviours toward sustainable retrofitting. More salient policy communications and guidance can contribute to increasing sustainability orientation and reducing perceived trade-offs with economic goals.

1. Introduction

1.1. Background and Research Question

The construction sector has a significantly large ecological footprint, accounting for substantial resource consumption and pollution [1]. In the European Union (EU), more than a third of the total consumed energy is used for building operations, with half of that energy used for room conditioning [2]. In 2020, Germany’s construction sector consumed half of all raw materials, created 60% of total waste, and generated 229.3 million tons of construction and demolition waste [3]. Conventional building materials, including concrete and steel, still dominate the industry, leading to significant CO₂ emissions. The cement sector and steel production alone contributed about 10% of global greenhouse gas emissions in 2020, a figure likely to grow with continued population increases and rising construction demands [4]. These trends amount to considerable pressure on the sector to contribute to the transformation and readiness for climatic change in built environments, particularly with respect to its vast building stock [5].

Progress in that direction is ambiguous, as evidenced by Germany’s retrofit rate being at just 0.72%, which remains well below the minimum annual requirement of 2% to achieve the national climate targets in the building sector by 2030 [6]. Previously explored drivers and barriers to sustainable constructions reveal a myriad of challenges, ranging from technical/engineering to organizational, political, and social challenges [7]. Specifically, not only is the technical complexity of retrofitting existing structures in dense urban areas complicated by increased standards and numerous norms [8], but also, retrofitting is growing into a social issue around housing accessibility and affordability [9,10]. The effects of the financialization of real estate are becoming increasingly apparent, with the affordable housing crisis as its most problematic symptom, which many cities around the world are experiencing [11]. Political countermeasures such as rent control or Milieuschutz (social preservation) zoning fail to address growing fears of green gentrification, highlighting the real or perceived tensions between environmental goals and social equity [12,13]. These challenges are compounded by the fragmented and adaptive nature of the construction sector, as well as its various stakeholders and their individual perceptions [14]. As Fernandez-Solis [15] argued, the construction process behaves as a complex system, highly sensitive to the early attitudes and priorities of its stakeholders. Adabre et al. [16] posited that sustainable and transformative strategies are first and foremost socio-technical, made possible by its actors, who must demonstrate behavioural efforts and willingness to adopt (partially) new practices. Additionally, previous studies in New Zealand and Canada show that stakeholders shape project sustainability through divergent understandings of what “sustainability” entails, generating tensions that affect the processes, resource allocation, fundamental planning decisions, and final outcomes [17,18,19]. Abdelaal and Guo [17] highlighted the interplay of knowledge, attitude, and practice as factors driving the adoption of green building certifications and noted that professional differences can stem from previous exposure and training. Specifically, without strong advocacy, the use of sustainable construction materials or circular procurement practices may be foregone by the project client in favour of saving costs [17]. Instead, the economic return and saving costs, despite potential social or ecological gains, are prioritized [14].

Their observations and findings suggest heterogeneity among practitioners in how barriers are perceived and managed and how industry readiness for sustainable transformation is appraised. According to Ajzen’s Theory of Planned Behaviour [20], attitudes and perceived control can shape intention and advocacy for sustainability inside organizations. Similarly, Bandura [21] emphasizes that repeated exposure and practice build efficacy, mastery, and motivation toward a practice, which may influence how challenges and one’s own behaviour are evaluated over time.

To better formulate policies and programmes for the German context that support socially appropriate, sustainable retrofitting, more empirical evidence regarding the specific challenges experienced by construction professionals is needed [22]. In this study, perceptions of the barriers and contextual conditions serve as early signals of readiness for action: they shape whether professionals feel able and willing to promote or champion sustainable practices. Because perceptions also reflect how favourable or “ready” the surrounding context feels (concerning regulations, material availability, and support mechanism), they offer indirect insight into conditions for sustainability.

The guiding research question is as follows:

How do construction professionals perceive key challenges to sustainable retrofitting, and to what extent are these perceptions shaped by environmental attitudes, age, profession, and sustainability engagement?

This question will be examined in the context of sustainable retrofitting in privately owned multi-storey residential buildings, a segment often underrepresented in favour of public-sector-focused sustainability research. In this way, contributions to a more differentiated understanding can be made regarding the main obstacles and contradictions of sustainable retrofitting and how policy needs to be reformed to realign environmental aspiration in a socially responsible manner.

This study reviews the relevant literature and empirically identifies key challenges to adopting sustainable retrofitting practices [23]. Those challenges are aggregated into political, organizational, and social barriers, which are then subject to rating by industry professionals, specifically architects, engineers, planners, and builders (developers). Based on a quantitative study with 86 participants from expert stakeholders in the construction industry, this study explores how construction professionals evaluate challenges such as legal uncertainty, funding gaps, skills shortages, and material constraints. Although developers often control the financing of construction projects, other professionals, such as architects, planners, and site managers, play a central role in how decisions are finally interpreted and implemented. Their expertise, priorities, and personal environmental attitudes may weigh in on the communication with clients and stakeholders to negotiate whether and how sustainability could be pursued in practice [15]. At the same time, their options are often limited by what is technically feasible, readily accessible, and affordable and must remain in line with construction regulations and norms.

The next subsections will briefly describe the relevant barriers and constructs of this study before outlining the applied methodology. After the presentation of the most relevant results, this study closes with a discussion and recommendations for policy and practice.

1.2. Sustainable Construction and Retrofitting in the German Built Environment

According to the German Federal Ministry for Housing, Urban Development and Building, sustainable construction can be described, much like the concept of sustainability itself, in terms of its three dimensions: economic, social, and ecological sustainability [24]. The economic dimension is concerned with factors such as construction costs, life cycle costs, cost-effectiveness, and value stability of building measures. The social dimension emphasizes human well-being, including health, participation, quality of life, esthetics, functionality, and the ability to meet basic needs. The ecological dimension stresses minimizing environmental impact, conserving resources, reducing land use, and enhancing biodiversity. Specifically, this entails reducing the building’s footprint; repurposing existing materials and structures; optimizing structural design to lower energy consumption during operation; incorporating non-toxic, durable, and easily separable materials; and minimizing water consumption [24]. Sustainable construction also requires fulfilling legal and performance-based requirements such as fire protection, sound insulation, and moisture protection at all times [24]. The social dimension of sustainability holds particular relevance for retrofitting. At its core, sustainable retrofitting is an expression of social responsibility toward future generations. The need to upgrade the existing building stock is driven not only by environmental goals but also by the ethical motivation to ensure that homes remain safe, liveable, and adaptable in the changing climate. Preparing buildings today helps avoid a future where the financial and social costs fall disproportionately on those who come after us. At the same time, retrofitting and the pursuit of decarbonizing cities is not without its caveats. As Rice et al. [25], and Miller and Mössner [26] critically argue, such spatially defined efforts can have unintended consequences, such as contributing to gentrification and displacement in the specific area, but more broadly, these measures can increase overall emissions through lifestyle-related consumption and commuting, creating possible rebound effects [8,27]. Their findings are a reminder to not overly rely on localized and technocratic solutions for greening the built environment but instead call for holistic and regional planning that anticipates trade-offs and considers social equity. Therefore, sustainable retrofit efforts must also respond to the needs of current populations, especially in light of rising housing costs and social inequality. Hence, retrofitting sits at the intersection of long-term responsibility and present-day justice and aligns the sustainability imperative with the human right to adequate housing [9,28,29].

Passer et al. [30] categorized retrofitting strategies into four broad categories: (1) reducing heating and cooling demand (e.g., insulation); (2) installing energy-efficient equipment and low-energy technologies; (3) integrating renewable energy systems and electrical upgrades; and (4) changes in human factors, including behaviour. This typology not only reflects both the technical and operational dimensions of retrofitting but also underscores that human behaviour is already recognized as a relevant element of sustainable building performance, even though it is rarely prioritized in implementation. Retrofitting fits into broader aspirations toward transitioning into a circular economy (CE), an economic system based on a closed-loop cycle of reusing, refurbishing, remanufacturing, and recycling materials [28,31]. CE is considered superior to the linear economy as it promotes resource sustainability, reduces waste, and encourages continuous material reuse [32] by taking into account future alternative uses of the space, easy disassembly, and repurposing, from which future generations can benefit. However, the uptake of previously used construction materials is rather low as safety and liability uncertainties remain high in a heavily regulated sector. Therefore, sustainable building efforts in Germany are still largely limited to operational energy efficiency in buildings in line with increasing policy mandates. These include measures subsidized by the Bundesamt für Wirtschaft und Ausfuhrkontrolle (BAFA) or the Kreditanstalt für Wiederaufbau (KfW), such as insulating building envelopes, replacing windows, and upgrading heating systems [33].

1.3. Organizational, Political, and Social Challenges in Retrofitting

The literature increasingly emphasizes that barriers to sustainable construction are not solely technical or financial but also include political, organizational, and social dimensions [34,35,36]. This study builds on these insights by focusing on how such challenges are perceived by practitioners in the field, who fulfil essential roles in the implementation of sustainable practices. In particular, this study is concerned with retrofitting of privately owned multi-storey residential buildings, a segment that is often underrepresented in research. At the organizational level, common barriers include skill shortages, both in regard to labour and expertise, high coordination costs, rigid cost structures, and market inertia for new practices and materials [37]. Retrofitting projects, with their inherent planning and technical complexities, make the need for digital and sustainable tools particularly apparent. Increased adoption of solutions such as Building Information Modelling (BIM) or material passports could help reduce coordination burdens, improve precision, and lower costs [38,39]. However, these tools are still rarely used in practice, as many firms lack the internal capacity or external partnerships to effectively implement them [38]. High upfront investment in sustainable planning can deter especially risk-averse developers, while warranty and liability concerns further discourage experimentation with reused or unconventional materials [27,37,40]. In addition, a lack of trust in the availability and quality of sustainable alternatives remains a common theme in both the literature and practice [41]. The fragmentation of the construction sector adds to the complexity. Each construction project is often a unique ensemble of practitioners, materials, and on-site conditions different from other manufacturing industries [15]. Developers, architects, contractors, and clients often operate with diverging priorities, leading to misalignment and possible blame-shifting [17].

Politically, inconsistent regulatory frameworks and limited financial incentives continue to hinder uptake. While energy-efficiency retrofits are eligible for subsidies, material innovation, lifecycle optimization, or truly circular retrofitting approaches receive limited support [42]. Complex applications and rules limit popularity. Moreover, outdated technical norms and protective regulatory structures can prevent the approval of recycled or non-traditional materials, reinforcing reliance on standard solutions [43,44]. This highlights the dilemma of political support as being both a driver and a barrier to the diffusion of sustainable construction.

Derived social challenges also play a role, particularly in cities where tenant protection laws and Milieuschutz zoning (socially protected areas) are in place to prevent displacement [45]. While such regulations are designed to support housing justice, they are sometimes perceived to limit financial returns on modernization, especially by developers, creating a real or perceived tension between social and environmental goals [13]. At the same time, fears of green gentrification and social backlash highlight the delicate balance between improving building performance and maintaining housing affordability [12].

1.4. Linking Structural Challenges to Individual-Level Attitudes and Behaviours

According to Fernandez-Solis [15], construction is marked by a “chaotic sensitivity to initial conditions” (p. 44), meaning that even small differences in stakeholder attitudes, for example, toward cost, quality, or sustainability, can result in widely different project paths, such as what is considered and what is abandoned. Preferences often change over time, adapting in response to practical constraints and arising dynamics during the planning, execution, and building phases. Because leadership is distributed and cooperation is rather ad hoc and situational, early stakeholder value-standpoints can play a disproportionate role in shaping retrofit outcomes [15]. While these barriers are well-documented at the policy and firm levels, less is known about how they are perceived by the actors implementing or advising retrofit projects on the ground. However, these perceptions matter. They shape decision-making; risk tolerance; and ultimately, the feasibility of advancing sustainability practices in real-world settings. This study examines how such barrier perceptions interact with four key individual-level variables—environmental attitude, age, professional role, and engagement profile—as reflected in their self-reported usage and perceived ease of sustainable practices. Each of these variables has been discussed in the literature as potentially shaping how sustainability challenges are perceived in practice [46,47].

With increasing age, practitioners may bring much industry experience and long-term exposure to changing regulations and practices. On the one hand, this familiarity with status quo methods may reinforce path dependency or scepticism towards new or sustainable-coined approaches [48]. On the other, continued encounters with regulatory hurdles and volatile markets may contribute to some level of frustration or fatigue that may amplify their perception of barriers. Age, therefore, may not only capture the accumulation of knowledge but also serve as a form of professional inertia and disillusionment that affects how challenges are evaluated in practice [40]. However, experience alone does not uniformly predict resistance; those with direct exposure to sustainable projects often report fewer perceived barriers, possibly due to increased confidence or positive reinforcement through practice [49]. Professional role reflects the respondent’s position within the construction process, such as architect, planner, site manager, or developer, and influences their exposure to specific constraints. For example, architects may focus on design and regulatory feasibility, while site managers are concerned with efficient on-site execution and handling time pressures. Prior studies have shown that these differing perspectives often lead to divergent views on the risks and practicality of sustainable construction [50].

Environmental attitude refers to the extent to which a professional supports or identifies with sustainability goals. Several studies have shown that stronger pro-sustainability attitudes are associated with greater openness to new approaches and lower perceived barriers, as individuals are more likely to see challenges as solvable rather than obstructive [46,47,51]. However, a positive attitude alone does not always lead to actual implementation in the presence of contextual hurdles. To account for this possible mismatch, this study also includes engagement profiles as a measure reflecting a professional’s familiarity with and possible mastery of sustainability methods, which might impact how barriers are perceived. Engagement is reflected in individuals’ reports of use of sustainable construction practices. Prior research suggests that professionals with more hands-on experience or frequent use of sustainable methods may feel more confident in managing related challenges and perceive fewer barriers [21,23,49].

Table 1. Challenge groups and short descriptions of the items.

Challenge Group	Challenge Items
Political Challenges	Lack of legal requirements, outdated or restrictive DIN standards, lack of government funding, rental price regulations, and regulations in social conservation areas.
Organizational Challenges	Lack of skilled labour, higher project acquisition and planning costs, product and services warranty and liability concerns, limited sustainable solutions, lack of information on materials and application, technical unsuitability of sustainable materials, other stakeholders’ lack of interest and awareness of sustainable practices, and absence of collaboration between stakeholders.
Social Challenges 1	Rental price regulations and regulations in social conservation areas.

¹ Social challenges were separated from policy challenges and tested for more nuance.

summarizes the relevant challenges that were subjected to stakeholder rating in this study.

2. Materials and Methods

2.1. Data Collection

This study administered a structured online survey to gather insights from construction professionals in Germany [52]. The questionnaire was developed in Google Forms and distributed via email and professional networks during September and October 2024. The survey was designed to minimize response fatigue by limiting completion time to under 10 min and included a total of 25 structured questions covering demographic information, and attitudes and experiences related to sustainable construction. The survey was conducted in German but subsequently translated into English for reporting.

To avoid confusion regarding terminology, a definition of sustainable construction was placed at the beginning of the questionnaire, in line with BMWSB definitions.

Participants then self-reported their use of sustainable practices, environmental attitude, and perceptions of specific challenges. An open-ended question invited participants to report further perceived barriers not captured in the predefined list. The questionnaire collected no other person-specific data other than age, gender, and profession and were kept anonymous. The survey did not collect any personal or sensitive data and posed no psychological or legal risks to participants. Participation was voluntary and anonymous, and this study complies with general ethical guidelines for social research. As the research did not involve vulnerable groups or trauma-inducing content, formal ethics board approval was not required.

2.2. Study Population and Sampling

This study focused on professionals operating in Germany’s private construction sector, including developers, architects, engineers, project planners, and site managers. Participants included those with experience in sustainable construction, as well as those who had not yet implemented such practices. Public-sector and cooperative housing projects were excluded due to their different financing and support conditions.

Given the lack of a centralized database of construction professionals in Germany, a non-probability sampling strategy was used, combining purposive and convenience sampling methods. Purposive sampling targeted professionals with relevant experience in retrofit or sustainable construction, in line with the aim of collecting information-rich cases [53]. To broaden the respondent pool and avoid bias toward sustainability-committed actors, convenience sampling was employed based on availability and accessibility [54].

Participants were identified via professional contacts within a network of practicing architects and developers, and through internet research using LinkedIn and Google. Individuals and companies were selected based on visible involvement in retrofit projects in the private sector. Snowball sampling was also applied by encouraging initial respondents to share the survey within their networks.

Out of 139 professionals contacted, a total of 88 responses were received during September and October 2024. After excluding two cases that did not meet the inclusion criteria (i.e., no experience in residential retrofit), the final sample consisted of 86 valid responses. The unit of analysis is the individual practitioner, rather than the organization or project level outcomes.

While this sampling strategy enabled us to reach professionals active in retrofitting, it also introduces the risk of over-representing sustainability-oriented respondents. The findings are therefore non-generalizable but offer theory-informed, experience-based insights into perceptions from within the industry. As a sensitivity analysis, we re-estimated our primary regression model after excluding the high-engagement subgroup (n = 63); the results remained substantively unchanged.

2.3. Key Variables and Their Operationalization

The goal of this study was to understand how professionals in the construction sector perceive key barriers to sustainable retrofitting and how these perceptions may be shaped by individual-level characteristics, such as their professional role, values, experience, and engagement.

2.3.1. Perceived Challenges to Sustainable Retrofitting

Challenges were conceptualized as dependent variables and were operationalized through 11 items that capture common barriers encountered in the planning and implementation of sustainable retrofitting projects. In the actual survey, respondents were asked to rate their agreement with each item on a 5-point Likert scale, ranging from 1 = strongly agree to 5 = strongly disagree. These items were adapted from earlier studies on sustainability barriers in construction [34,35,36,37,50,51,55].

The items were then grouped into two internally consistent categories. Internal reliability was assessed by calculating Cronbach’s α. The first group represents policy challenges (5 items; Cronbach’s α = 0.79), addressing issues such as lack of financial incentives, insufficient legal clarity, and restrictive norms and regulations. The second group captures organizational shortfalls (6 items; Cronbach’s α = 0.93), including perceived skill shortages, elevated planning costs, and limited access to suitable materials.

In addition, two exploratory items were grouped together to reflect social dimensions of retrofit resistance: rent control (Mietendeckel) and socially protected areas (Milieuschutz). The two-item social scale showed high internal consistency (α = 0.96) and broadens the scope of perceived challenges to include housing-related considerations.

2.3.2. Environmental Attitude

Environmental attitude was captured via a three-item composite measure (5-point Likert) and checked with Cronbach’s α. Respondents were asked to rate (1) how important the implementation of sustainable practices is for reducing CO₂ emissions and conserving resources, (2) whether they prefer to use environmentally friendly materials in their projects, and (3) their personal motivation to incorporate sustainable practices and materials in the future. All items were rated on 5-point Likert scales (original: 1 = strongly agree, 5 = strongly disagree). The internal consistency of this scale was validated using Cronbach’s α = 0.92 (95% CI: 0.885–0.945), indicating that the items reliably capture a coherent underlying concept. This operationalization reflects both a dimension of normative commitment and behavioural intention, aligning with Ajzen’s (1991) theory of planned behaviour, and is consistent with previous studies that emphasize the role of attitudinal readiness in sustainability adoption [46,47,51].

2.3.3. Professional Background

Participants self-reported their profession: builder/developer, project planner, architect/engineer, or site manager. The underlying rationale for including this variable is that certain professional responsibilities result in more likely confrontations with certain challenges, as well as certain barriers and their interpretations. For example, builders/developers may emphasize leveraging public funding and keeping an eye on the return on investment, whereas architects often engage more with design constraints and are responsible for ensuring compliance with current regulations and specifications. Site managers, in turn, are positioned at the interface between design and on-site execution and are more likely to assess the feasibility of sustainable measures and possible trade-offs under real-world conditions. They may prioritize previous routinized processes over new approaches, which require more effort to adopt. Prior research has emphasized that differing backgrounds and resulting orientations can lead to diverging perceptions of risks and opportunities [50].

2.3.4. Engagement Profile

To approximate behavioural mastery/exposure with sustainable retrofitting, participants were asked to indicate which sustainable practices they had implemented in their previous work. Based on their responses, they were grouped into one of several engagement profiles. These range from low/no engagement to categories such as circular economy adopters or high-engagement users.

Table 2. User profiles with coding logic and group collapse.

Profile ID	Collapsed Subgroup	Label	Coding Logic
1	High	High Engagement	Selected ≥ 6 distinct measures
2	Medium	CE-Oriented	Selected ≥ 3 from CE_DfD, CE_Reuse, CE_Recycle, CE_Reduce
3	Medium	Bio and Energy Adopter	Selected ≥ 1 Bio and Energy = 1, total categories = 2–3
4	Medium	Material-Focused	Selected ≥ 3 Material_Bio, not in other groups
5	Low	Energy Only	Energy = 1, and total selected = 1
6	Low	Tech-Enabled	Digital = 1, total selected = 1–2, not in other groups
7	Low	Low/No Engagement	Selected ≤ 1 measure or “none”
8	Medium	Other	No profile matches measures ticked

Note: High = ≥ 6 distinct measures; low = ≤ 1 measure; medium = all others.

below provides an overview of the possible profiles and their assignment criteria.

The engagement profiles thus reflect both readiness and familiarity with various kinds of sustainability approaches. Earlier findings suggest that those who have implemented even a limited number of sustainable measures tend to perceive fewer institutional and technical barriers [40,49]. Although not part of the engagement profile itself, usage frequency and ease of implementation were also determined using a Likert-scale from 1 to 5. Together, these indicators reflect the practitioner’s familiarity and growing sense of competency with sustainable retrofit practices.

These underlying concepts are also reflected in Ajzen’s theory of planned behaviour, where repeated behaviour (e.g., usage of sustainable measures) strengthens perceived control and might reinforce similar future behaviour. While the engagement profile reflects some real-world behaviour, it may underestimate the actual involvement of professionals who planned or coordinated measures without executing them directly. For this reason, engagement profiles are understood here as a proxy for experiential familiarity and prior exposure.

The goal of creating the profiles was to obtain a qualitative sense of what measures are being considered by professionals and to monitor the performance of important concepts such as circular economy or technology-aided tools, as they remain issues themselves, aside from retrofitting, and represent a bit more novelty compared with using wood elements. Therefore, some distinct subgroups, such as the CE-oriented, tech-enabled, and material-focused subgroups, were created. These group labels were used to provide descriptive insights. However, for the inferential analyses, we later collapsed these groups into subgroups (high, n = 23; medium, n = 31; and low, n = 32) to ensure sufficient cell sizes.

2.3.5. Age Group

Although age is often treated as a control variable, it is included in this study as a potentially meaningful factor influencing how professionals perceive challenges. It may reflect generational shifts in professional training, as well as differences in the degree of exposure to updated sustainability standards, but also implementation hurdles. In this context, age is expected to shape how challenges are perceived as either more manageable or increasingly inhibiting, depending on prior and role-specific experience [40]. The survey collected data on four age groups, which were kept for descriptives and group comparisons (Kruskal–Wallis followed by Dunn’s test), but for inferential analysis, we collapsed to two broader groups into one (younger 18–40 and older (41–56+)).

2.4. Data Preparation and Analysis Approach

The analysis combined descriptive and inferential techniques to explore how individual characteristics shape their perceptions of retrofit-related challenges. The survey data was exported to excel, and where necessary, the data was cleaned and Likert-scale items reverse-coded to ensure comparability and improve interpretability. This step allowed all scores to follow a consistent direction, whereby higher values indicate stronger agreement with the presence of a challenge or, in the case of implementation questions, greater ease. Through this coding strategy, mean scores, correlations, and regression coefficients could be interpreted uniformly and with less ambiguity.

The data was cleaned using Excel. Then, the data tables and keys for the numerical answers, as well as their underlying concepts, were converted into CVS files and placed into the data folder. All further analyses were conducted in the Jupyter notebook using Python 3.12, which ran in the development environment set up in Visual Studio Code. All analyses were run locally in python using open-source libraries such as pandas, numpy, scipy, statsmodels, scikit-posthocs, python-docx, and matplotlib. Linguistic flow, analytical logic, visual checks, and interpretive support were iteratively refined with the assistance of ChatGPT (Version 4o and 5) to ensure code quality and clear table outputs. ChatGPT did not perform any statistical estimations or produce any results. The results presented here are reproducible from Jupyter notebook.

The descriptive analysis included frequencies, means, standard deviations, and visualizations such as boxplots and histograms to identify trends and outliers.

Given the ordinal structure of many survey items that used the Likert-scale ratings of perceived challenges, and the possibility of response clustering (e.g., a general tendency to agree or disagree across challenges), we decided to use Spearman’s rank correlation to explore associations among key variables. Spearman’s p does not assume linear relationships, which makes it more appropriate for detecting non-linear trends in datasets with categorical or ordinal variables, which are present in this study.

To explore group-level differences in perceived challenges and attitudes, particularly across engagement profiles and professions, non-parametric Kruskal–Wallis tests were performed [56], followed by Dunn’s post hoc tests to identify significant pairwise differences [57].

Ordinary Least Squares (OLS) regressions were used to further explore whether perceived challenges relate to the key independent variables: age, attitude, and engagement level. Our dependent variables were composite indices (averages of multiple Likert items) for organizational, policy, and social challenges. Treating such indices as continuous is common in survey research [58,59] but because this choice can be debated, several diagnostic and robustness checks were carried out [60]. To avoid unstable estimates, we used the collapsed age groups and the three collapsed engagement profile groups. The pairwise Spearman correlation matrix showed that attitude and engagement profile are moderately correlated (Spearman p ≈ −0.37), which raised concerns about possible collinearity and destabilization of the regression model. All models were checked for key assumptions: Breusch–Pagan tests showed no signs of heteroskedasticity, and multicollinearity was within acceptable ranges once profiles were excluded.

To further strengthen confidence in the results, we conducted several robustness checks. First, we re-estimated the models without the high-engagement subgroup to reduce potential sampling bias (n = 63), which produced very similar patterns. Second, we ran ordered-logit regressions on tercile-split outcomes to ensure that the findings were not dependent on continuous Likert indices. Summaries of the diagnostic model can be found in

Table A1. Other models—key statistics (no coefficients).

Outcome	Model	n	Adj. R2	ΔAdj. R2 vs. M1	AIC	BIC	BP p	Normality p	Max VIF	Profile Block F p
Organizational	M2 (M1 + C(profile3))	86	0.709	−0.006	150.5	165.2	0.494	0.63	4.02	0.789
Organizational	M3 (M1 excl. High)	63	0.708	−0.007	107.0	115.6	0.281	0.849	4.92
Organizational	OL1 (Ordered Logit, full)	86			164.1	181.3
Organizational	OL3 (Ordered Logit, excl. High)	63			117.5	132.5
Policy	M2 (M1 + C(profile3))	86	0.566	0.002	152.9	167.7	0.804	0.068	4.02	0.371
Policy	M3 (M1 excl. High)	63	0.588	0.025	106.7	115.3	0.409	0.027	4.92
Policy	OL1 (Ordered Logit, full)	86			164.3	179.1
Policy	OL3 (Ordered Logit, excl. High)	63			120.4	133.2
Social	M2 (M1 + C(profile3))	86	0.649	−0.002	210.6	225.4	0.455	0.003	4.02	0.564
Social	M3 (M1 excl. High)	63	0.664	0.013	139.8	148.3	0.394	0.0	4.92
Social	OL1 (Ordered Logit, full)	86			178.2	195.4
Social	OL3 (Ordered Logit, excl. High)	63			118.9	133.9

Note: OLS uses HC3-robust SEs; ordered logit uses default asymptotic SEs. BP: Breusch–Pagan (H₀: homoskedastic). VIFs < 5 in all OLS models. Δ values are relative to Model 1.

.

3. Results

3.1. Descriptive Overview of the Sample

The descriptive statistics shown in

Table 3. Participant demographics and survey descriptions.

Gender	Count	%	Profession		Count	%
Female Male	45 41	52.3 47.7	Architects/Engineers Project Planners Builders/Developers Site Managers		33 21 17 15	38.4 24.4 19.8 17.4
Age Group	Count	%	User Profile		Count	%
18–29 30–39 40–55 56+	15 21 27 23	17.4 24.4 31.4 26.7	1—High Engagement 7—Low/Now Engagement 3—Bio and Energy Adopter 8—Other 5—Energy Only		23 22 18 13 10	26.7 25.6 20.9 15.1 11.6
Variables		Mode	Median	Mean	Std. Dev.	Min.	Max.
Organizational Challenges Policy Challenges Social Challenges Sustainable Retrofit Attitude Ease of Implementation Measure Usage Frequency Measure Count		3.88 5.00 5.00 5.00 2.00 2.00 1.00	3.56 4.00 4.00 4.00 2.00 2.00 2.00	3.34 3.86 3.50 3.84 2.41 2.69 3.37	1.04 0.86 1.34 1.03 1.01 1.14 3.13	1.00 1.67 1.00 1.67 1.00 1.00 0.00	5.00 5.00 5.00 5.00 5.00 5.00 17.00

Note: n = 86. Minimum and maximum values reflect observed respondent-level averages for multi-item scales. 1 = low, 5 = high.

provide a glimpse of the characteristics of the sample. The sample had a fairly balanced gender distribution (52.3% female; 47.7% male). The respondents represented a broad age range, with the majority between 40 and 49 years (31.4%), followed by those aged 50–59 (26.7%), and only 17.4% were under the age of 30. Given their professions, this age distribution is somewhat expected as many of these roles require a certain level of experience and know-how. Regarding professional background, architects and engineers comprised the largest subgroup (38.4%), followed by project planners (24.4%), builders/developers (19.8%), and site managers (17.4%).

3.2. Use of Sustainable Retrofitting Measures

The reported number of implemented practices ranged widely across the sample (M = 3.37, SD = 3.13, min = 0, max = 17), reflecting notable variation in exposure, practice, and familiarity. The survey also sought to categorize participants according to these measures to create profiles reflecting their engagement with sustainable retrofitting measures. Out of the eight constructed engagement profiles, only five were actually represented in the sample. The largest share belonged to the “high-engagement” group (26.7%), indicating a broad-based application of various practices (a minimum of six unique measures were ticked). A comparable share (25.6%) fell into the opposite category of “low/no engagement”. Other profiles, such as “Bio & Energy Adopters” (20.9%) or “Energy Only” users (11.6%), reflect more siloed forms of engagement. Although not all profiles were eventually represented, the coding for specific types of measures allowed for qualitative insights into what commonly known practices are applied in the retrofitting arena of this specific sample.

Table 4. Sustainability measures and their application distribution.

ID	Short Description	Code (Used to Create Profiles)	Count	Selected (%)
M1	Switch to sustainable heating (e.g., heat pumps and solar)	Energy	56	65.1
M2	Use of modular/pre-fab components	CE_DfD (Design for Disassembly)	6	7
M3	Use of deconstructable building parts/materials	CE_DfD	2	2.3
M4	Use of rain/grey water for irrigation/toilets	NBS (Nature-based solutions)	25	29.1
M5	Replace taps with water-saving models	CE_Reduce	10	11.6
M6	Reuse of previously used construction material	CE_Reuse	10	11.6
M7	Reuse of previous construction elements	CE_Reuse	9	10.5
M8	Façade and roof greening	NBS	28	32.6
M9	Use of BIM software	Digital	4	4.7
M10	Use of material passports	Digital	1	1.2
M11	Involving renters during planning	Social	6	7
M12	Use of recycled concrete	CE_Recycle	15	17.4
M13	Use of wood for façades, ceilings, and roofs	Material_Bio	31	36
M14	Use of wood-hybrid structures	Material_Bio	14	16.3
M15	Use of clay plasters or bricks	Material_Bio	5	5.8
M16	Straw or wood-based wall construction	Material_Bio	18	20.9
M17	Natural floor coverings (linoleum, cork)	Material_Bio	54	62.8
M18	None of the above	None	9	10.5

presents the number and type of measures and how many respondents indicated their previous experience.

A more detailed breakdown shows that within the low/no-engagement group, nine participants reported that they had not implemented a single predefined sustainability measure and did not use the self-report box to disclose alternative measures. The remaining 13 respondents indicated experience with only one category (not energy), typically limited to wood use or natural floor coverings. Furthermore, tenant participation, which is often framed as an important driver in reconciling social-related tensions in retrofitting policy, was only reported by six respondents; notably, five of whom belonged to the high-engagement group. This suggests that tenant participation remains an exception rather than the norm in construction projects. Similar goes for the limited adoption of digital tools: only four individuals reported using digital tools such as BIM or material passports, which points to the continued marginality of digital tools for retrofitting, at least within the surveyed sample. As for professional groups, the boxplots show no notable differences in the uptake of measures.

3.3. Challenge Perception and Environmental Attitude

Table 5. Ranked challenge ratings.

Challenge	Challenge ID and Category	Median	Mean	Std. Dev.
Higher costs for sustainable construction, incl. planning	C8, Organizational	5	4.38	0.90
Insufficient public funding opportunities	C4, Policy	5	4.36	0.84
Lack of awareness and interest in sustainable construction of other stakeholders	C13, Organizational	4	4.19	1.03
Lack of tax relief	C1, Policy	4	4.09	0.98
Insufficient regulations to apply sustainable construction practices and material	C2, Policy	4.5	4.02	1.22
Outdated and restrictive DIN norms and regulations	C3, Policy	4	3.69	1.12
Rent price regulations, such as rent cap	C5, Policy, Social	4	3.57	1.32
Regulations to protect social conservation areas (Miliueschutz)	C6, Policy, Social	4	3.43	1.43
Lack of providers for sustainable materials and services	C10, Organizational	4	3.31	1.43
Lack of skilled labour for sustainable construction	C7, Organizational	3.5	3.24	1.43
Lack of warranties for sustainable material and service provision	C9, Organizational	5	3.17	1.36
Technical unsuitability of sustainable construction material and practices	C12, Organizational	3	2.98	1.39
Lack of access to information about sustainable materials and their application	C11, Organizational	3	2.95	1.48
Absence of collaboration between stakeholders	C14, Organizational	2	2.48	1.29

provides a ranking of the challenges that were subject to rating by the participants. Among the top eight challenges, five of them were perceived as highly challenging, and half of them are related to policy instruments. Additionally, half (C8, C4, C1, and C5) were associated with economic concerns around profitability and financial support.

With regard to the remaining variables, the descriptive statistics for the three challenge groups, environmental attitude, usage frequency, and perceived ease of implementation provide further orientation. All challenge categories received moderately high mean values, which suggests that they are indeed experienced as considerable hurdles in practice, which is consistent with their conceptual framing as barriers to sustainable retrofitting. The policy-related challenges had the highest mean and agreement (M = 3.86, SD = 0.86), while the social dimension was most divergently rated (M = 3.50, SD = 1.34) by the practitioners in this sample, indicating that there might be some factor driving the varying perceptions of this specific challenge group. The average environmental attitude enjoyed a relatively high rating among the entire sample (M = 3.84, SD = 1.03), indicating an overall alignment with pro-sustainability values. Furthermore, the perceived ease of implementation (M = 2.41) and self-reported usage frequency (M = 2.69) suggest moderate familiarity with sustainable measures as well as a slight notion of difficulty in executing retrofit projects.

3.4. Spearman Correlation

The descriptive statistics already provided useful insights into the structure and tendencies within the sampled data, revealing some possible patterns and relationships. To examine how the variables relate to each other more systematically, a Spearman rank-order correlation was performed using Python 3.12 statistical packages in vs. Code, the results of which are presented in

Table 6. Correlation table.

	1	2	3	4	5	6	7	8	9	10	11
1. Attitude
2. Profile	−0.366 **
3. Profession	0.076	−0.210 †
4. Gender	−0.207 †	0.183	0.118
5. Age	−0.539 ***	0.187	−0.122	0.193
6. Ease	0.512 ***	−0.341 **	0.076	−0.002	−0.172
7. Measure_count	0.559 ***	−0.614 ***	0.176	−0.111	−0.152	0.604 ***
8. Usage_freq	0.535 ***	−0.416 ***	−0.016	−0.073	−0.087	0.742 ***	0.759 ***
9. Org_avg	−0.747 ***	0.231 †	−0.085	0.139	0.634 ***	−0.493 ***	−0.417 ***	−0.421 ***
10. Pol_avg	−0.657 ***	0.233 †	−0.062	−0.007	0.517 ***	−0.459 ***	−0.519 ***	−0.492 ***	0.796 ***
11. Soc_avg	−0.610 ***	0.350 **	−0.289 *	0.090	0.613 ***	−0.249 *	−0.410 ***	−0.264 *	0.743 ***	0.841 ***

Note: Significance levels: † p < 0.10; * p < 0.05; ** p < 0.01; *** p < 0.001. n = 86.

. As expected, all challenges are highly correlated with each other, which is a logical outcome, given that all items were theoretically derived and all serve as conceptual barriers to sustainable retrofitting. Notably, age and attitude are significantly and negatively correlated (−0.54 ***), whereas attitude, measure count, ease of implementation, and usage frequency are all significantly and positively correlated. Since the measures were used to create the engagement profiles, not surprisingly, they have an almost perfect correlation (0.95 ***), which indicates an appropriate categorization of the profiles. These patterns indicate overlap between profiles and behavioural exposure, but not perfect substitutes. Accordingly, we retained attitude and age in the regression models and evaluated the incremental value of profile terms via model comparison rather than including them by default. Furthermore, unlike profession, which was only significantly negatively correlated with social challenges (−0.29 **), profile was very highly and significantly correlated with all three perceived challenges.

3.5. Group Comparisons with Kruskal–Wallis Omnibus and Post Hoc Dunn–Holm

Given the Likert-type indices and group nature of the data exceeding two levels, we used the Kruskal–Wallis (KW) omnibus test to compare differences in outcome medians [58], followed by Dunn’s post hoc tests with Holm step-down correction for pairwise comparisons when the omnibus was significant [57,58]. Alongside p-values (denoted with asterisks (*)), we report effect size as epsilon-squared values (ε²_H).

3.5.1. Profession

The results reveal a statistically significant difference in the perception of social challenges across professional groups (H-test = 9.092, p < 0.0281 *, ε²_H = 0.07). Dunn’s test identified that developers/builders rated social challenges significantly higher than project planners. For reference, builders (n = 17) had a median rating of 5.0 (very high), architects (n = 33) had a median of 4.0, project planners (n = 21) had a median of 3.0, and site managers (n = 15) had a median rating of 3.5. No professional differences were significant for organizational and policy challenges.

3.5.2. Profile

To reduce errors and increase cell size for more robustness, we used the collapsed subgroups for the KW test and compared the high (high engagement), medium (all others, with min. two measures), and low (low/no engagement and energy only) groups. We report consistent differences for all three challenge types, which are also presented in

Table 7. Kruskal–Wallis omnibus tests for subgroup differences (age, engagement profile, and profession) across perceived challenge indices and attitude. Entries report H(df) and ε²_H. * denotes Holm-adjusted significance for the omnibus test. Higher scores indicate more perceived barriers.

	Outcome	Organizational Challenges	Policy Challenges	Social Challenges	Attitude
Factor
Age (18–29, 30–39, 40–55, 56+)		H(3) = 54.63 *, ε2_H = 0.61 **	H(3) = 46.14 *, ε2_H = 0.51 **	H(3) = 40.57 *, ε2_H = 0.44 **	H(3) = 44.19 *, ε2_H = 0.49 **
Profile (High, Low, Medium)		H(2) = 13.19 **, ε2_H = 0.13	H(2) = 17.22 *, ε2_H = 0.18	H(2) = 13.10 **, ε2_H = 0.13	H(2) = 23.46 *, ε2_H = 0.25
Profession (1. Builders/Developers, 2. Architects/Engineers, 3. Project Planners, 4. Site Managers)		H(3) = 4.78, n.s.	H(3) = 0.35, n.s.	H(3) = 9.09 *, ε2_H = 0.07	H(3) = 6.76, n.s.

Note: significance levels: ** = p < 0.01; * = p < 0.05; n.s. = not significant (p ≥ 0.05). Age uses four bands (18–29; 30–39; 40–55; and 56+). Engagement uses collapsed subgroup profiles (high; low; and medium). Profession comprises four role families.

:

• Organizational challenges: H(2) = 13.19, p = 0.001, ε²_H = 0.13;
• Policy challenges: H(2) = 17.22, p < 0.001, ε²_H = 0.18;
• Social challenges: H(2) = 13.10, p = 0.001, ε²_H = 0.13.

The medians of the profiles follow the same pattern, where high-engagement groups report perceiving the fewest challenges and those in the low-engagement group report the most, e.g., organizational medians: high = 2.75, medium = 3.25, and low= 4.00. These findings indicate that professionals with more exposure and familiarity with sustainable practices tend to perceive retrofit barriers as less severe. While this does not imply causality, the results reflect Bandura’s [1] notion that repeated experiences can reinforce confidence and shape how challenges are appraised.

3.6. Regression Results

Following the descriptive statistics, correlation patterns, and group comparisons, individual-level variables such as age, attitude, and engagement profile were further tested in regression models [61,62], not intended as a predictive model but as an exploratory tool to examine the associations between perceptions of challenges and individual characteristics more systematically. The results of the OLS regression of Model 1 are presented in

Table 8. Model 1 OLS moderation results.

Model 1. Outcome: Organizational Challenges				Model 1. Outcome: Social Challenges
Variables	β (HC3)	SE (HC3)	95% CI	Variables	β (HC3)	SE (HC3)	95% CI
Intercept (Age 18–40)	2.961 ***	0.193	[2.583, 3.338]	Intercept (Age 18–40)	3.557 ***	0.246	[3.074, 4.039]
Attitude (mean-centred)	−0.725 ***	0.201	[−1.119, −0.332]	Attitude (mean-centred)	−1.471 ***	0.238	[−1.939, −1.004]
Age: 41–56+ (vs. 18–40)	0.896 ***	0.221	[0.463, 1.329]	Age: 41–56+ (vs. 18–40)	0.744 **	0.28	[0.196, 1.292]
Interaction (Attitude × Age 41–56+)	0.42	0.22	[−0.011, 0.851]	Interaction (Attitude × Age 41–56+)	1.438 ***	0.264	[0.920, 1.955]
Adj. R2 = 0.715, BP p = 0.33, Normality p = 0.603, Max VIF = 3.11, n = 86				Adj. R2 = 0.651, BP p = 0.511, Normality p = 0.0 Max VIF = 3.11, n = 86
Model 1. Outcome: Policy Challenges
Variables	β (HC3)	SE (HC3)	95% CI
Intercept (Age 18–40)	3.655 ***	0.164	[3.335, 3.976]
Attitude (mean-centred)	−0.624 ***	0.152	[−0.923, −0.326]
Age: 41–56+ (vs. 18–40)	0.606 **	0.2	[0.214, 0.997]
Interaction (Attitude × Age 41–56+)	0.432 *	0.176	[0.088, 0.777]
Adj. R2 = 0.563, BP p = 0.432, Normality p = 0.07, Max VIF = 3.11, n = 86

Note: n = 86, significance level: *** p < 0.001, ** p < 0.01, * p < 0.05. Coefficients use HC3 SEs; 95% CIs in brackets. Attitude is mean-centred. SEs are HC3-robust. Normality p = Shapiro–Wilk on residuals; Max VIFs < 5.

.

Main Model: Interaction Analysis with Age Group and Attitude

Model 1 specification:

challenge = β₀ + β₁ attitude (mean_c) + β₂ age_older + β₃ (attitude (mean_c) × age_older) + ε.

We estimated OLS moderation models with attitude (mean-centred) [58,63], age groups (41–56+ vs. 18–40), and their interaction (Model 1; n = 86). The results across all challenge groups are consistent: respondents with higher pro-environment attitudes reported perceiving fewer challenges, while older respondents perceived challenges more strongly. The attitude × age interaction term indicated that this negative attitude–barrier link was weaker among older respondents. Attitude did not have the same buffering effect on challenge perception for older professionals as it did for younger professionals. This was most apparent for policy and social challenges and was only directionally for organizational challenges.

The fit for Model 1 was reasonable to strong across outcomes, and basic diagnostics were acceptable. The applied Breusch–Pagan tests did not indicate heteroskedasticity, and multicollinearity stayed within common guidelines in Model 1 (all VIFs < 5). Because residual normality was not perfect, especially for the social outcome (Shapiro–Wilk test for social challenges < 0.001 (not normal), we reported HC3-robust standard errors throughout to provide more conservative confidence intervals.

We ran two robustness checks by adding engagement profile as a categorical block to Model 2 (M2, Table A1). Adding these profiles did not improve the explanatory power or raise the VIFs. To address possible sampling and sustainability self-selection bias, we excluded “high-engagement” respondents (M3, n = 63), thus producing very similar outputs: attitude is still associated with reporting lower challenge perception and older respondents with higher perceptions, and the moderation pattern remains. Ordered-logit models (treating outcomes as 5-level ordinal, OL1 and OL3) reproduced the same signs for the key associations.

However, some notable limitations arose during the diagnostics and robustness checks, especially in the OL models. Removing the high-engagement group from an already modest sample created a slight imbalance by reducing the younger subsample more than the older one (13 out of 23 high-engagement respondents belonged to the younger group). This made some ordinal categories thin in the OL3 subset and widened a few CI. After rounding to 1–5, no cases fell into category “1” for policy challenges (the lowest mean score in the data was above 1.5). Additionally, social challenges showed clear non-normal residuals in OLS; using HC3 was therefore important. Both oddities reflect a clear limitation of the survey design. As the outcomes were challenges, challenge perceptions were expected to gather at the higher end, with rather few cases where these shortcomings were not perceived as a barrier.

4. Discussion

4.1. Discussion of the Results

This study sought to investigate how construction professionals perceive key challenges to sustainable retrofitting and how these perceptions are shaped by environmental attitudes, age, profession, and engagement profiles. Although the sample (n = 86) is modest and the results cannot be generalized, the insights from the non-random expert sample offer valuable insights into a diverse and typically hard-to-reach population. The results show that environmental attitudes were high across the sample (M = 3.84), suggesting that value-wise professionals recognize the need for sustainability in the industry. At the same time, all three challenge domains (policy, organizational, and social challenges) received moderately high ratings, confirming that well-known barriers such as regulatory complexity, institutional inertia, and material availability remain problematic [35,37]. The relatively high environmental attitude observed in this study is at odds with the high challenge ratings of political support and perceived low interest in sustainable construction by other stakeholders. Recent work by Liu et al. [64] suggests that older adults in China may be more receptive to policy cues if they are salient, thus resulting in translations into environmentally friendly behaviour. These findings bring up an important issue with policy communication and concerns that sustainability in construction continues to be a systemic issue. In the context of the German construction industry, this could mean that it is not a lack of individual motivation but an issue of supportive conditions and lack of policy salience under which professionals operate [64,65]. The rather low uptake of sustainable retrofitting measures by around three quarters of the sampled experts suggests a mismatch between market and policy incentives, and personal motivation and readiness.

Professional background was not as important to the challenge ratings as expected. Only developers rated social challenges significantly higher than planners, possibly reflecting concern over tenant regulations such as rent caps or Milieuschutz zones that make investments risky or impossible [12]. These same policies were rated far less problematic by younger respondents, hinting at a generational divide. One interpretation is that younger professionals, many of whom may be affected by high rent prices themselves, are more sympathetic to such protections and may not perceive, or may not want to perceive, them as insurmountable barriers to retrofitting, remaining hopeful for the compatibility of environmental and social goals. Older professionals, by contrast, may associate calls for more sustainability in construction with yet another wave of regulations that result in long approval periods and longer sales cycles while also demanding more efforts due to possible learning curves for new practices and material applications.

4.2. Limitations and Future Research

A key limitation of this study lies in the sampling strategy and the sample size (n = 86), which does not support population inference. Due to the selective and snowball sampling, there is a chance of self-selection toward sustainability-oriented respondents, which further limits generalizability of the findings and should therefore be interpreted as associational patterns within this sample. The differences we observed may not generalize to all construction professionals in Germany or elsewhere. Furthermore, the data is based on self-reported perceptions, which might not reflect actual behaviour. To mitigate and stabilize the results stemming from our data, we collapsed sparse categories (e.g., age 18–39 vs. 40+, Profile 3) to stabilize estimates and reported effect sizes with 95% CIs.

Future studies should plan for the use of larger samples and, where feasible, use stratified recruitment across roles and regions across Germany. Our outcomes for this study are perceived organizational, policy, and social challenges, and the perceptions in this study are conceptualized as proximal indices to the contextual factors around a possible sustainability transformation in the industry, in this regard, indirectly collecting insights into how favourable or ready practitioners perceive their environment to be for sustainable practices/transformation. Still, perceptions and engagement, as captured in this study, do not explain actual behaviour. Attitude–behaviour gaps are well documented in sustainability research [63]. Our results suggest that closing this gap likely depends on what sits between attitude and action in practice. Based on the present patterns (age differences, some engagement effects, and the moderation by attitude), future studies could more specifically test constructs such as self-efficacy and perceived behavioural control regarding contextual factors and retrofit measures, similar to a Li et al. [66]. Future research can triangulate individual-level characteristics and contexts with actual project sustainability metrics, such as LCA comparisons of retrofits designed with circular design and recycled products vs. conventional approaches, LCA-based cost analysis of energy retrofits and grey-energy payback time, energy-class improvements, and comparisons of rent process before and after retrofit measures.

4.3. Implications and Recommendations

To address the concerns of older cohorts, low engagement, and specifically social challenges, several practical recommendations can be made to developers/builders. Rising construction costs make these findings plausible. Rent protection laws are designed to prevent luxury retrofits that trigger gentrification [9,12]. At the same time, they can extend to longer payback periods for investors and are associated with lower and longer periods of return on investment. To avoid gentrification while still encouraging necessary retrofitting to maintain building quality and reduce carbon emissions, early tenant involvement is crucial. Research shows that including tenants can reduce tensions between renters and developers [15,19]. Moreover, Alshukri et al. [67] found that actively involving stakeholders in processes enhances internal capabilities through learning and innovation, which strengthens sustainable value creation. Simple steps, such as stakeholder mapping and seeking honest dialogue, may improve the onboarding and willingness of renters to pay possible increased rents if they can expect tangible benefits, such as lower energy bills and indoor air quality. Routine exchange and collaboration with occupants can also prevent unnecessary and costly retrofit measures by encouraging energy conservation behaviours after retrofit measures [68]. Mutual understanding reminds developers that they are providing housing, not only managing a financial asset. Public recognition for developers who seek tenant dialogue and maintain affordable rent prices could further encourage such practices. A possible explanation for older cohorts reporting high perceptions of barriers is regulation fatigue and role overload, especially for those in senior roles with higher responsibility. Tax breaks should be given to SMEs and single entrepreneurs who adopt digital tools to encourage uptake and free funds to hire extra personnel to lighten the load and help navigate administrative tasks.

Training programmes are another important driver in overcoming constraints and hesitancy toward sustainable retrofit measures and maintain positive attitudes, with continuing and adult education allowing for sufficient time to develop both digital and practical skills. In Germany, the Bildungsurlaub scheme, including education vouchers, offers a good entry point but is underutilized [69]. At present, participation often depends on individual initiative and employers’ willingness to subsidize costs. To overcome these barriers, a stronger public commitment regarding fully subsidizing programmes targeting the acquisition of sustainable construction skills and stakeholder management, with financing provided through government sponsorship, is needed. A curated portfolio of certified training modules that cover energy-efficient construction, circular economy practices, introduction to bio materials, and digital tools such as BIM [38,39], would increase exposure among practitioners who are otherwise constrained by time and financial pressures. Such early and repeated exposure builds both technical mastery and self-efficacy (cf. Bandura, 1982 [21]), which in turn reshapes professional attitudes toward sustainability.

While encouraging continued education is crucial in the long-run, policy support must become more attentive to time constraints and the friction during real adoption in the construction industry. In line with Liu et al. [64], policies need to be communicated more saliently, with information dissemination directed to professionals via social media possibly making information more accessible and fostering understanding and self-efficacy.

Sustainability-oriented tools such as BIM, digital building passports, or life-cycle-based tendering are rarely used [38,39], which may not be attributable to the professionals’ lack of awareness but rather their lack of capacity to learn about and sustainably implement them when pressing daily project deliverables demand full attention.

In line with other authors’ recommendations, not only retrofitting but also construction, in general, deserves more holistic planning, i.e., in the case of cultural heritage buildings, to consider both the cultural value and energy efficiency [8]. The digitalization in support systems needs to streamline application processes for obtaining public funding or permits for retrofitting and needs to be made more accessible and user-friendly. Moreover, more and better online access is needed for sustainable materials that do not require excessive search times or long and expensive wait times regarding approval for serial production. These shortfalls, however, also represent major business opportunities to support the industry in its transformation [70]. Furthermore, as other studies have stressed before, public-sector projects can legitimize sustainable retrofits, but they should ensure public acceptance and clearly demonstrate community benefits, given the increased sensitivity in housing inequality and the need for more access to shared spaces [12,27,29].

The industry urgently needs increased awareness among political representatives about the multiple pressures construction professionals are facing. Beyond decarbonization, these professionals are expected to deliver housing affordability, and lasting and circular quality [71], while complying with stringent regulations and evolving standards, which must often happen under uncertain contracts and unstable skilled labour supply [72]. As Fernandez-Soliz [15] reminds us, sustainability in the built environment requires asking not just how we build, but why, so that necessary adaptations and green solutions that ultimately can reconcile long-term planetary needs with present human realities can become the norm [73]. While some challenges to sustainable retrofitting around the availability of solutions and reducing the complexity of building regulations remain, most persistent challenges touch upon the affordability of retrofitting, which affects both investors and renters. Current policies to curb rising costs for shelter in a highly financialized real-estate market are unintentionally contributing to the deadlock situation, where landlords avoid the cost of (sustainable) retrofitting if they cannot recoup the investment costs in a desired timeframe, and some tenants save on operational costs, while some tenants do not gain any benefits from retrofit savings in the face of higher rent costs [14,22].

5. Conclusions

This study suggests that, even among a motivated expert group, perceptions of barriers remain high, while the uptake of sustainable measures is comparatively low. Although this study’s results cannot be generalized due to the modest sample size, the observed patterns highlight meaningful associations: strong environmental attitudes are linked to lower perceived barriers, yet this buffering effect weakens for more seasoned professionals. Developers, in turn, tend to perceive social challenges such as tenant protections as particularly discouraging to sustainable retrofitting.

These findings point to broader structural conditions in the industry, where sustainable approaches still struggle to gain mainstream acceptance despite climate targets and funding schemes promoting energy retrofitting. They also underline that construction must not be framed only as an emitter to be regulated but also as a central intersection of societal and economic organization, and a crucial driver of adapting to climate change. Sustainability necessitates considering all facets and taking a holistic approach, which call for continued education and exposure to sustainable practices and the active involvement and management of stakeholders [22].

Practitioners are advised to take note of the value in early tenant involvement, routine communication with stakeholders, and investment in digital and sustainable tools as strategies that can reduce perceived barriers and strengthen retrofitting acceptance.

Policy responses must address sustainability as both housing and social policy issues, not merely a technical fix to environmental problems. This means reducing the bureaucratic overhead and creating more accessible and targeted funding schemes for SMEs and cooperatives. Investing in continued education fosters practical exposure and a pro-environmental mindset and levels the ground for the industry’s readiness for a sustainable transition [74].

Finally, continuing attempts to reconcile all three sustainability dimensions are not evidence of failure of the concept itself but indicate the need to reimagine the purpose of the built environment and a political recommitment to housing as a common good that is designed and maintained for current and future generations who will live in it.