A Corporate Social Responsibility (CSR) Model to Achieve Sustainable Business Performance (SBP) of SMEs in the South African Construction Industry

Abstract

This paper purposes to develop a corporate social responsibility (CSR) model to guide small and medium enterprises (SMEs) in the South African construction industry (SACI) towards sustainable business performance (SBP). A theoretical CSR model was developed from the literature and validated through Partial Least Squares Structural Equation Modelling (PLS-SEM), using primary questionnaire data obtained from 110 SMEs in the SACI, who are registered on the construction industry development board (cidb) register of contractors between Grade 1 and 6 general building (GB) and/or civil engineering (CE). The PLS-SEM results indicate that CSR drivers influencing the CSR practices of SMEs, along with CSR implementation challenges experienced by SMEs, influence the perception of SMEs pertaining to the relationship between the integration of CSR and SBP all of which influences the CSR activities considered by SMEs to achieve SBP, thus implying that the holistic adaption of the PLS-SEM (CSR model) by SMEs in the SACI equates to more SBP. However, SMEs practicing CSR activities to achieve SBP are subjected to specific CSR: drivers and implementation challenges as well as SME owner perceptions.

1. Introduction

A thriving economy depends primarily on the capacity to establish an excellent business environment for small and medium enterprises (SMEs) that supplies quality services and competitive products at affordable costs and in quantities that are market related [1]. Thus, it is important for SMEs to acclimatise to fluctuating conditions of competition and innovation through the process of globalisation [1,2]. In South Africa, SMEs have been identified by the government as a source of job creation to boost the economy. The role of SMEs in the SACI should therefore be considered as important to the economy as the role of SMEs in general; [3], cited by [4,5], therefore assert that construction SMEs are critical drivers of economies locally and globally.

It is well acknowledged that South Africa’s small business sector is experiencing many challenges. One challenge is that the birth of SMEs in South Africa, according to Total Early-Stage Entrepreneurial Activity (TEA), is of the lowest in the world [6]. Despite this, South Africa’s TEA ranking has improved to 27 out of 54 countries [7], with the small business sector continuing its contribution towards the South African economy. According to Bisseker [8], South African SMEs employ 47% of the workforce, contribute to more than 20% of GDP and pay about 6% of all corporate taxes. These are good statistics in terms of the contributions of SMEs towards society but unfortunately many SMEs show signs of unsustainable business performance in their first year of existence [9]. Vallie [10] reports that a staggering 70–80% of SMEs in South Africa do not manage to survive the first year of business. These signs of unsustainable business performance are not only related to SMEs in general but also reflect negatively on construction SMEs which form part of the South African business sector at large. This is justified by Vallie [11] who reports that statistics currently illustrate that 70–80% of construction SMEs fail within their first five years of existence. This raises huge question marks around the sustainability of construction SMEs in South Africa. According to Vallie [11], construction SMEs in South Africa are presently battling to accomplish their growth potential to become pivotal drivers of job creation; in addition, South Africa’s sluggish economic growth, ongoing political uncertainty and current national budget shortfall of ZAR 209 billion, all contribute to the current state and position of construction SMEs across South Africa. To sum up, in recent years, the South African government has impacted negatively on SMEs by allowing the small business environment to deteriorate significantly.

To overcome this problem, authors such as Aigbavboa and Thwala [5], Ramukumba, Mofokeng, Eke et al., Aigbavboa et al., and Bushe [12,13,14,15,16] have suggested that the major contributor threatening the sustainable business performance (SBP) of SMEs in the SACI relates primarily to the lack of SME owner–managers’ management knowledge (planning capacity, resourcing capacity, leadership capacity and controlling capacity); business knowledge and self-knowledge; industry experience in the chosen area of business such as construction; and business acumen, aptitude and entrepreneurial mind-set to raise a successful enterprise. These studies, however, have not placed emphasis on corporate social responsibility (CSR) as a driver to the SBP of SMEs in the SACI, which is supported by [17,18]. The lack of emphasis on CSR as a driver agrees well with the findings of a study conducted by [19], which indicates that CSR research in the construction industry worldwide is still in its initial stages. Xia et al. [19] further confirm utilising descriptive methods such as frequency and percentage, and figures across the Scopus academic database, Emerald, Taylor and Francis, Elsevier and Google Scholar—that between 2005 and 2017, 68 reliable journal papers were sourced relating to CSR in the construction industry. These articles were sourced from 21 countries covering all the continents. From these 21 countries, the UK (19) and China (10) had the highest number of papers, followed by Australia (9), the USA (4), South Korea (4) and others (2 and below). Xia et al. [19] also mention that these journal papers covered themes relating to CSR perceptions, CSR dimensions, CSR implementation and CSR performance. By contrast, a study by [20] determined that, between 2006 and 2018, 50 journal papers and 19 conference papers could be sourced relating to the following CSR themes linked to construction enterprises: drivers of CSR implementation; motivation for CSR implementation; and barriers to CSR implementation. The 69 papers sourced by [20] included 17 countries, both developed and developing, with the majority of papers focusing on China and the UK. These statistics give a clear indication as to the limited CSR research in the construction industry globally.

This limitation also extends to South Africa where only eight articles [21,22,23,24,25,26,27,28] could be sourced relating to CSR and its implementation in the SACI. Of these, only three focused on SMEs. The first study [22] attempted to establish the status of small, medium and micro-enterprises (SMMEs) in the built environment in relation to CSR to promote an awareness of the CSR function in the community, promoting SME growth, improvement and sustainability. The second study [24] focused on establishing the extent to which construction SMEs in Gauteng, South Africa involve CSR in their practices. The third study [27] focused on addressing the negative trajectory of SMEs in the SACI by exploring the concept of CSR, investigating the organisational perceptions of SMEs relative to the relationship between the integration of CSR and SBP. Applying the studies conducted by [19,20] as a benchmark, it is evident that no significant study pertaining to the research themes highlighted by [19,20] have been conducted in South Africa, particularly from a construction business perspective, which includes SMEs operating in the SACI. Research relating to the use of a CSR model to guide SMEs in the SACI towards achieving SBP is also limited. Expanding on the knowledge presented by [19,20], this study aimed to develop a CSR model to guide SMEs in the SACI towards the achievement of sustainable business performance, focusing on the following CSR related factors and their hypothetical relationships:

• The CSR drivers that influence the CSR practices of SMEs in the SACI;
• The challenges that SMEs in the SACI experience pertaining to the implementation of CSR;
• The perception of SMEs in the SACI pertaining to the relationship between the integration of CSR and sustainable business performance;
• The CSR activities that must be considered by SMEs in the SACI to achieve sustainable business performance.

To achieve this a quantitative research approach in the form of an online questionnaire survey, to validate the CSR model through PLS-SEM, was used.

2. Theoretical and Conceptual Framework

2.1. Perspective of CSR Theories

All theories in CSR serve as a point of reference for every set of CSR practices, and, since there is no single accepted theory and perspective of CSR, there should be extensive variation in what constitutes the theoretical and practical aspects of CSR [19]. According to [20], the theories underpinning CSR studies express how CSR is observed or interpreted by different business stakeholders from different perspectives. This study, however, was guided by the definition of CSR by [29] who stated that CSR is considered as:

a concept whereby companies integrate social and environmental concerns in their business operations and in their interaction with their stakeholders on a voluntary basis.

For the purpose of this study, the following theories—instrumental theory, legitimacy theory, stakeholder theory and perception theory—were examined, of which one or a combination of these theories will support the conceptual framework and therefore the achievement of the research aim.

2.1.1. Instrumental Theory and Its Link to CSR

According to Garriga [30], with regards to instrumental theory, CSR is only a strategic tool to achieve economic objectives that include competitive advantage over competitors and ultimately wealth creation for a business. Hamidu et al. [31] concur with [30], stating that instrumental theory views CSR from the perspective of a strategist aiming to practice CSR as an indispensable opportunity to exploit and receive economic benefits for the business. Hamidu et al. [31] further explain that instrumental theory accentuates the association of CSR practices of a business with the maximisation of profits to benefit various business stakeholders. Representative of instrumental theory is the familiar view of [32]:

The only one responsibility of business towards society is the maximisation of profits to the shareholders within the legal framework and the ethical custom of the country.

Hence, when a business commits to CSR to support its core business activities, accompanied by substantial profit maximisation, the commitment assumes a strategic position in the decision-making process of the business if underpinned by instrumental theory [31]. Many studies in the general business environment which include [17,33,34,35,36] uphold instrumental theory and its link to CSR practice agreeing that there is a positive relationship between a business practicing CSR and its overall financial performance. Similarly, studies [23,37], in the construction business environment, share the same sentiments. This therefore serves as a justification that instrumental theory does encourage the commitment to CSR practices by businesses if it allows for profitability, competitive advantage, good business image or reputation enhancement.

2.1.2. Legitimacy Theory and Its Link to CSR

Legitimacy theory, like any other theory, is perceived by various authors from different perspectives. According to Suchman [38],

Legitimacy is a generalised perception or assumption that the actions of an entity are desirable, proper or appropriate within some socially constructed system of norms, values, beliefs and definitions.

Likewise, Lindblom [39] insists that legitimacy is a condition or status that exists when a business’s value system correlates with the value system of a major portion of a society. Based on morality, [40] views legitimacy as a process of justification by which a business strives to justify itself to its peers or to its superordinate systems of its right to exist. In addition, [41] support the view of [40], adding that a business is legitimate when judged to be just and worthy of support by society. In contrast, when an actual or potential disparity exists between the business and societal value systems, there is a threat to the entity’s legitimacy [31]. Thus, with regards to the legitimacy theory via CSR practices, a business, throughout its survival, needs to fulfil what the society expects from it and, by adhering to the society’s expectations, the business is considered deserving of being in the same environment with the society it serves. Again, the business is regarded as part of society with a legitimate right of survival. In other words, the legitimacy theory argues that a business can only continue to exist if the society in which it is based perceives the business to be operating with a value system (demonstrating CSR initiatives) that is commensurate with the society’s own value system [42].

According to Hamidu et al. [31], it is critical to note that legitimacy theory deals with two major concepts: the first is the way in which the general public perceives the business, and the second is the efficiency of the communication channels between the business and its society. Between the two concepts, the efficiency of the business communication channel to the surrounding society is most essential with regards to legitimacy theory and therefore as a link to practicing CSR through legitimacy theory; communicating the business CSR practices is advantageous, allowing for the initiation and protection of the business legitimacy. Hence, [43,44,45] all observe a positive relationship between the disclosure of CSR practices and a business’s legitimacy to survive in society.

2.1.3. Stakeholder Theory and Its Link to CSR

To understand stakeholder theory in its totality, defining who and what a stakeholder represents is very important. Freeman [46] suggests that a stakeholder is defined as:

Any group or individual who can affect or is affected by the achievement of the firm’s objectives.

Clarkson [47] defines stakeholders as any person or group of people who have an ownership right or any form of interest or claim to a business. Starik [48] includes human and non-human entities in the definition of a stakeholder, and by doing so regards the natural environment as the non-human stakeholder due to the implications and relevance it has on CSR policies. Clarkson [47] further classifies stakeholders into two groups, namely primary and secondary groups. According to Hamidu et al. [31], the primary group of stakeholders are those who have a direct impact on the survival of the business, as their continuous participation allows the business to survive. For example, the business is only able to survive if its managers utilise their skills by establishing valuable products to satisfy its shareholders, customers, suppliers, partners, investors, employees and government. According to Hamidu et al. [31], secondary stakeholders to a business do not directly impact the achievement of the business objectives as their roles and responsibilities are less important so business survival does not depend on whether or not they participate in the business. From a construction business perspective,

Table 1. Stakeholders and categories relevant to construction businesses worldwide (source: [49]).

Project Level Category	Business Level Category
Employees	Employees
	Shareholders
Customers
Suppliers and Partners
	Government
Environment
Community	Community
Competitors	Competitors
	NGOs

indicates according to Zhao et al. [49] two categories of stakeholders specifically mapped for construction enterprises, namely: project level stakeholders and business level stakeholders.

According to Mitchell et al. [50], stakeholder theory extends the business objectives from profit maximisation to the satisfaction of stakeholder needs and, despite the criticism that stakeholder theory receives, it is supported by empirical studies that indicate that many businesses partake in CSR to serve stakeholder demands [51]. Stakeholder theory as a link to CSR definitely specifies how a business—more specifically in the case of this study a construction business, bearing in mind the stakeholder categories as stipulated by [49]—should implement CSR without viewing CSR as an isolated concept [47]. Linking stakeholder theory to the implementation of CSR practices, as documented by [20], will lead to construction businesses, large or small, achieving business goals and subsequently SBP [17,18] concur that business performance is measured by the way it satisfies its stakeholders as there is a positive relationship between stakeholder satisfaction and SBP.

2.1.4. Perception Theory and Its Link to CSR

Gibson’s theory of direct perception alleges that the cognitive thinking of human beings evolved by the evolutionary influence of the external environment, causing the receptors of human beings to be sensitive to relevant stimuli in the environment, allowing for cognitive adaptation [52]. Demuth [52] states that this type of interpretation of perception is called ecological perception as the external environment plays a pivotal role in influencing the process of perception in human beings. According to Robbins [53], various factors within the external environment such as time of event, work setting, social setting and background influence the perception of human beings. In addition, [54] insists that past and present experiences (challenges, failure, success), assumptions and expectations, education, self-concept, culture, faith, values, preconceived notions and present circumstances all influence the perception of humans and thereby impact the decision-making process.

This information can also be correlated to the way in which business owners (human beings), particularly construction business owners such as SMEs, perceive their business environment and the way business decisions, particularly around CSR initiatives and activities, are made, considering CSR drivers and implementation challenges influence perceptions and choices of CSR practice [20]. Gibson’s other theory relating to perception in action posits that perception is viewed as a requisite property of animate action, arguing that without perception being realised, action would be unguided and, without action, perception would serve no purpose [55]. From a business perspective, [56] supports this view, stating that decision making is an important and necessary skill that a business owner needs to acquire in order for a business to achieve its goals and objectives. Furthermore, UK Essays [56] mention that organisational excellence, which encompasses SBP, leans heavily on proper decision making by the business owner and management team, guided by their perceptions. Consequences from these decisions can therefore make or break a business, so UK Essays [56] feel strongly that business owner perceptions and decision-making processes work hand in hand, stating the following:

Perception plays a vital role in the decision-making process. Therefore, decision makers often use perception to create, evaluate, and choose decision options.

2.2. Selection of the Theory Underpinning the Conceptual Framework and Research Aim

All four theories are linked to the practice of CSR. However, based on the research aim, it is important that the term SBP be acknowledged as a key driver in the selection of any one, or combination, of the aforementioned theories to support the conceptual framework and the achievement of the research aim. The term SBP is defined in this study as follows:

A phenomenon maintaining a set of attributes at a productive level or rate, focusing on the construction SMEs’ stakeholder satisfaction provided in a culture that motivates the construction SME owner–manager to coordinate construction resources and activities.

[27]

From this definition and for the purpose of this study, SMEs in the SACI should acknowledge and maintain their stakeholder satisfaction levels, which include not only the satisfaction of the society in which they operate, but also the satisfaction of their employees, shareholders, customers, suppliers and partners, government, the environment and resources, competitors and NGOs to achieve SBP. This is supported by [20] who suggest three distinct constructs influencing CSR practices (underpinned by stakeholder theory) of construction businesses globally: CSR drivers; motivation to practice CSR; and CSR barriers. These constructs and possible relationships amongst them are open to further debate and investigation. Thus, this is the point of departure for this study considering stakeholder theory and perception theory as the most relevant theories to support the conceptual framework outlined by Figure 1.

2.2.1. Variables

The conceptual framework of this study, as presented in Figure 1, illustrates the research, latent and measured variables that will be examined. The conceptual framework has four latent variables, namely: SME perceptions pertaining to the relationship between the integration of CSR and SBP; CSR drivers influencing CSR practices of SMEs; CSR implementation challenges experienced by SMEs; and CSR activities to be considered by SMEs to achieve SBP. According to Kenton [57], these latent variables cannot be directly measured and are assigned measured variables that have been established through the extensive review of literature. From the conceptual framework, it can be seen that the measured variables associated with the latent variables are:

• Measured Variables 1: internal and external organisational perceptions;
• Measured Variables 2: international CSR drivers and national CSR drivers;
• Measured Variables 3: normative management level challenges, strategic management level challenges, operational management level challenges and environmental management level challenges;
• Measured Variables 4: employee dimension activities, shareholder dimension activities, customer dimension activities, supplier and partner dimension activities, government dimension activities, environment and resources dimension activities, community dimension activities, competitor dimension activities and NGO dimension activities.

Figure 1. Conceptual framework (source: [58]).

Figure 1. Conceptual framework (source: [58]).

2.2.2. Conceptual Framework—Predicted Hypothetical Relationships

The conceptual framework (illustrated in Figure 1) predicts the following hypothetical relationships presented in

Table 2. Predicted hypothetical relationships between latent variables (source: [58]).

Hypothetical Relationships	Supporting Theory	Supporting Literature
Hypothesis 1: there is a significant relationship between CSR implementation challenges experienced by SMEs and SMEs’ perceptions pertaining to the relationship between the integration of CSR and SBP.	Perception theory	[53,54]
Hypothesis 2: a significant association exists between CSR drivers influencing the CSR practices of SMEs and SMEs’ perceptions pertaining to the relationship between the integration of CSR and SBP.	Perception theory	[53,54]
Hypothesis 3: there is a significant relationship between CSR implementation challenges experienced by SMEs and the CSR activities considered by SMEs to achieve SBP.	Perception theory and stakeholder theory	[20,31,59,60,61,62,63,64,65,66]
Hypothesis 4: a significant affiliation exists between SMEs’ perceptions pertaining to the relationship between the integration of CSR and SBP, and the CSR activities considered by SMEs to achieve SBP.	Perception theory and stakeholder theory	[19,23,36,37,67,68,69,70,71,72]
Hypothesis 5: there is a significant relationship between CSR drivers influencing the CSR practices of SMEs and the CSR activities considered by SMEs to achieve SBP.	Perception theory and stakeholder theory	[20,69,73,74,75,76,77,78,79,80,81]
Hypothesis 6: merging the CSR implementation challenges experienced by SMEs, and SMEs’ perceptions pertaining to the relationship between the integration of CSR and SBP, as well as CSR drivers influencing the CSR practices of SMEs, impacts the CSR activities considered by SMEs to achieve SBP.	Perception theory and stakeholder theory	Ref. [20] including a combination of all the other supporting literature.

, supported by theory and literature, that exist between the latent variables.

3. Methods

A mixed-method approach in the form of an explanatory sequential design was adopted for this study, utilising a structured questionnaire survey in phase 1 of the data collection process and structured interviews in phase 2. This approach was taken to investigate the latent and measurement variables of the study as indicated in Section 2.2.1. The justification for the choice of this design stems from the fact that the study required a large strand of quantitative data in phase 1 of the data collection process and a smaller strand of qualitative data in phase 2. This smaller strand of qualitative data collected in phase 2 was utilised to compile case studies that allowed for the sourcing of valuable information to complement the quantitative data with the aim of developing a comprehensive CSR model to guide SMEs in the SACI towards a sustainable business performance. Hence, the design of the structured interviews pertaining to phase 2 of the data collection process took place once the data pertaining to the structured questionnaires in phase 1 was collected and analysed.

The questionnaire survey and structured interview consisted of five sections each, with the questionnaire survey asking closed-ended questions and the structured interview open-ended questions. For the questionnaire survey, the first section required respondents to answer questions pertaining to demographical information, with the second, third, fourth and fifth sections requiring respondents to rate statements that describe the CSR drivers that influence their CSR practices; the challenges that their organisations experience pertaining to the implementation of CSR; their perception pertaining to the relationship between the integration of CSR and sustainable business performance; and the CSR activities that they consider to achieve sustainable business performance. Similar to the questionnaire survey, the structured interviews also required respondents to answer general demographical questions. Thereafter, open-ended questions regarding their perception pertaining to the relationship between the integration of CSR and sustainable business performance; the drivers influencing their CSR practices; the CSR implementation challenges they experience; and the CSR activities that they consider to achieve sustainable business performance were asked.

3.1. Sample and Data Collection (Questionnaire Survey—Phase 1)

The sample survey participants for the online questionnaire survey were drawn from the national cidb register of contractors, who occupy General Building (GB) and Civil Engineering (CE) classes of work between Grade 1 and Grade 6 (Table 3), and who according to Windapo et al. [82] represent the SME cluster in the SACI, with current upper limit tender values, ranging between less than ZAR 1 million to ZAR 20 million. To determine a suitable representative sample for this study, the formula by [83], cited in [84,85], was applied:

$$ss=z^{2}x\frac{p\left(1-p\right)}{C^2}$$

(1)

where

ss = sample size

z = standardised variable

p = percentage picking a choice, expressed as a decimal

C = confidence interval, expressed as a decimal

To achieve a sample size with a given degree of accuracy, the worst-case percentage picking choice of 50% was assumed [84,85,86]; a 95% confidence level was assumed as in other studies with a significance level of α = 0.05; z = 1.96 at 95% confidence level; and a confidence interval (c) of ±10%.

Table 3. Research target population (source: [27,58]).

Province	Western Cape	Northern Cape	North West	Mpumalanga	Limpopo	Kwa Zulu Natal	Gauteng	Free State	Eastern Cape	Total
Cidb Grade
1 GB	1828	804	2910	2802	3024	6901	11,154	1481	2819	33,723
1 CE	413	270	972	1489	1553	5289	1863	842	2500	15,191
2 GB	77	25	70	47	43	89	234	29	60	674
2 CE	24	15	19	51	27	58	66	33	117	410
3 GB	27	3	11	10	12	31	58	7	19	178
3 CE	22	4	5	25	14	163	31	24	27	315
4 GB	51	2	15	18	28	37	83	5	25	264
4 CE	37	2	11	22	28	78	55	22	36	291
5 GB	20	3	0	6	9	15	30	2	4	89
5 CE	19	2	8	18	17	51	19	10	13	157
6 GB	29	2	6	17	14	28	32	5	13	146
6 CE	25	7	5	20	19	49	47	8	18	198
Total	2572	1139	4032	4525	4788	12,789	13,672	2468	5651	51,636

Table 3. Research target population (source: [27,58]).

Province	Western Cape	Northern Cape	North West	Mpumalanga	Limpopo	Kwa Zulu Natal	Gauteng	Free State	Eastern Cape	Total
Cidb Grade
1 GB	1828	804	2910	2802	3024	6901	11,154	1481	2819	33,723
1 CE	413	270	972	1489	1553	5289	1863	842	2500	15,191
2 GB	77	25	70	47	43	89	234	29	60	674
2 CE	24	15	19	51	27	58	66	33	117	410
3 GB	27	3	11	10	12	31	58	7	19	178
3 CE	22	4	5	25	14	163	31	24	27	315
4 GB	51	2	15	18	28	37	83	5	25	264
4 CE	37	2	11	22	28	78	55	22	36	291
5 GB	20	3	0	6	9	15	30	2	4	89
5 CE	19	2	8	18	17	51	19	10	13	157
6 GB	29	2	6	17	14	28	32	5	13	146
6 CE	25	7	5	20	19	49	47	8	18	198
Total	2572	1139	4032	4525	4788	12,789	13,672	2468	5651	51,636

The sample size was computed as follows: $ss=\frac{1.962 \times 0.5 \left(1 - 0.5\right)}{0.12}=96.04$.

According to Takim et al. [87] the response rate is usually in a range of 20–30%. Consequently, it was necessary to adjust the sample size to account for non-responses. Assuming a conservative response rate of 20%, the appropriate sample size to be surveyed was calculated as follows:

$$Survey ss=\frac{new ss}{response rate}$$

(2)

$$Survey ss=\frac{96}{0.2}=480$$

(3)

The survey sample size was therefore approximately 480 respondents, who were randomly selected. Active contact details were obtained from the cidb, for the online questionnaire survey. A cover letter including the link to the online questionnaire survey (LimeSurvey) as seen in Appendix A was sent out via email to all research participants (

Table 4. Sample survey participants (source: [27,58]).

Province	Western Cape	Northern Cape	North West	Mpumalanga	Limpopo	Kwa Zulu Natal	Gauteng	Free State	Eastern Cape	Total
Cidb Grade
1 GB	16	6	6	6	6	16	16	6	6	84
1 CE	16	6	6	6	6	16	16	6	6	84
2 GB	6	5	5	5	5	6	6	5	5	48
2 CE	6	5	5	5	5	6	6	5	5	48
3 GB	5	2	2	2	2	5	5	2	2	27
3 CE	5	2	2	2	2	5	5	2	2	27
4 GB	5	2	2	2	2	5	5	2	2	27
4 CE	5	2	2	2	2	5	5	2	2	27
5 GB	5	2	2	2	2	5	5	2	2	27
5 CE	5	2	2	2	2	5	5	2	2	27
6 GB	5	2	2	2	2	5	5	2	2	27
6 CE	5	2	2	2	2	5	5	2	2	27
Total	84	38	38	38	38	84	84	38	38	480

). One week after the last cover letters and questionnaires were sent out, 38% of the targeted sample size indicated their willingness to partake in the study. Participants who did not respond were contacted via telephone calls as a follow-up. Further to the follow-up calls, 71% of the targeted sample size agreed to participate; however, this did not guarantee the rate of questionnaire completion and submission.

The internet-mediated questionnaire survey approach was used to reach a large audience throughout various provincial regions with a wide geographical dispersion. How-ever, some of the respondents’ emails bounced back, while some respondents opted out, based on reasons that they were not interested in operating in the construction sector any longer; others simply opted out because they were too busy, among other things prioritizing their business around the COVID-19 pandemic, and were unable to attend to the questionnaire. Most of these reasons were received and noted via the follow-up telephone calls that were made to the respondents. To achieve a high response rate from the participants who showed interest to participate in the survey, notifications requesting their response to the questionnaire survey were sent on a weekly basis to enhance their interest for the research and to ensure a good response rate. Of the 480 questionnaires surveys emailed to respondents, 110 were suitably completed and returned, resulting in a response rate of 23%. The questionnaire distribution compared to the responses received for the different provinces is reported in

Table 5. Questionnaire distribution vs. responses received (source: [27,58]).

Province	Distributed	Received	Percentage
Western Cape	84	51	60.7
Northern Cape	38	4	10.5
North West	38	2	5.3
Mpumalanga	38	7	18.4
Limpopo	38	2	5.3
Kwa Zulu Natal	84	11	13.1
Gauteng	84	18	21.4
Free State	38	3	7.9
Eastern Cape	38	12	31.6
Total	480	110	22.9

.

3.2. Sample and Data Collection (Structured Interview—Phase 2)

To eliminate bias in the selection of participants to be interviewed a formal letter (see Appendix B) was sent via email on 17 May 2021, asking construction organisations who partook in the first phase of the data collection process (quantitative data collection—Section 3.1) if they would consent to be interviewed for the purpose of achieving the overall objectives of this research. Respondents who indicated a willingness to be interviewed were immediately contacted via email requesting an interview appointment date, which was then scheduled by the researcher (as indicated in Appendix C). From the 110 responses received at the close of the quantitative survey, four respondents with head offices in Cape Town and smaller offices in other parts of South Africa indicated interest in participating in the interview. All four respondents were contacted and individual interview dates scheduled with each participant, taking into consideration postponements that might arise due to the COVID-19 pandemic. The first interview was thus conducted with the owner of Organisation A (cidb Grade 6 GB and CE) on 27 May 2021, at 16:00 at the organisation’s head office in Cape Town, with the interview recording lasting 58 min. The second interview was conducted with the owner of Organisation B (cidb Grade 4 GB and CE) on 28 May 2021, at 15:00 at a neutral venue (coffee shop) in the northern suburbs of Cape Town, lasting 45 min. The third interview was conducted with the owner of Organisation C (cidb Grade 1 GB and CE) on 31 May 2021, at 18:00 at a neutral venue (coffee shop) in the southern suburbs of Cape Town, lasting 30 min. The fourth interview was conducted with the owner of Organisation D (cidb Grade 2 GB and CE) on 3 June 2021, at 19:00 at a neutral venue (coffee shop) in the southern suburbs of Cape Town, lasting 63 min. The researcher requested permission from all four interviewees to use a digital voice recorder to record the interview. Permission was granted to the researcher.

3.3. Method of Data Analysis

Data analysis was carried out in three parts. The first set of data presented and analysed is the demographical information pertaining to the research participants, followed by the extraction and presentation of the measurement variables linked to each latent variable as seen in Appendix D, compiled from the data collection and analysis regarding the questionnaire survey and structure face-to-face interviews. Thereafter, the measurement and structural models are analysed. The analysis of the demographical information was conducted using descriptive (mean, standard deviation, etc.) statistics. The analysis of the measurement model started by drawing all possible structural relationships between the latent variables of the study, allowing for the reflective indicators of the latent variables to turn from red to blue, indicating some form of relationship with each other. Thereafter, the PLS algorithm determined the standardised regression rate, factor loadings and the percentage variance R-squared (R²) value explained by the explanatory variables. This study considered 0.5 as the baseline for factor loading, as acceptable [88]. The analysis of the measurement model thus also tested the convergent and discriminant validity.

The structural model was analysed, by running the PLS algorithm to identify the relationship (if any) existing among the variables. The PLS algorithm was run to identify the variance explained by the variables included in the model and to establish the significance levels of the paths leading to the PLS estimate. The path coefficients were also evaluated to identify the contributions of each latent explanatory construct to the predictive capacity of the endogenous construct. The overall predictive capacity of the structural model, according to Chin [89], was also assessed by the R² value associated with the endogenous constructs within the model. To establish the significance level of the variables, the bootstrapping technique was performed using 500 resamples. This illustrated the structural model with path coefficients and t-statistics. The assumption with regards to bootstrapping, more specifically the t-statistics, is that a t-statistic above 1.65 indicates that the path coefficient is significant at p ≤ 0.10. If the t-statistic is greater than 1.96, the path coefficient is significant at the p ≤ 0.05 significance level and, when the t-statistic is above 2.57, it is significant at p ≤ 0.01 [90]. Considering the ongoing need to report and evaluate the performance of PLS models, including both measurement and structural models, and with attention to the overall predictive power of the model, a global criterion of goodness of fit (GoF) index as recommended by [91] was used. The procedural guidelines provided by [92] to compute the GoF values, which are minimum values for global validation of PLS path models, were followed.

4. Results and Discussion

4.1. Construction Related Experience and Education of SME Owner

Examining the data in

Table 6. Years of construction related experience pertaining to SME owners (source: [27,58]).

Variable		Frequency	Percentage	Cumulative Percentage
Years of construction related experience pertaining to SME owners	1–5 yrs	43	39.1	39.1
	6–10 yrs	25	22.7	61.8
	11–15 yrs	25	22.7	84.5
	16–20 yrs	8	7.3	91.8
	21–25 yrs	9	8.2	100
	Total	110	100

it is clear that of the SME owners who partook in the questionnaire survey 39% have obtained between 1 and 5 years of construction related experience; 23% have acquired between 6 and 10 years of construction related experience; another 23% have obtained between 11 and 15 years of experience; with just 7% acquiring between 16 and 20 years of construction related experience; leaving 8% of SME owners se-curing between 21 and 25 years of construction related experience. Further analysing the data, it is clear that approximately 61% of the SME owners have an average of 16 years of construction related experience, indicating significant experience in the construction sec-tor, contributing to the validity and reliability of their responses.

The data in

Table 7. Construction related education and training of SME owners (source: [27,58]).

Variable		Approximate Percentage
Construction related (university) education and training of SME owner:	ND Civil Engineering	15%
	ND Building	9%
	BTech Construction Management	6%
	BTech Civil Engineering	4%
	BTech Quantity Surveying	3%
	BSc Construction Management	2%
	BSc Civil Engineering	1%
	ND Architecture	1%
	MSc Project Management (Construction)	1%
Construction related (TVET) education and training of SME owner:	Qualified Bricklayer	16%
	Cert. Building and Civil Engineering	3%
	Qualified Tiler	2%
	Qualified Carpenter	2%
	Qualified Plasterer	1%
	Qualified Plumber	1%
	Cert. Construction Project Management	1%
	Cert. Road Construction	1%
	Cert. Construction Maintenance	1%
	Cert. Contractor Development Programme	1%
Other education and training of SME owner:	Qualifications in: Business Management; Education; Internet Technology; and Mechanical Engineering	23%
	Matric Certificate	3%
SME owner with no qualification:	None	3%

indicates that roughly 71% of the SME owners have obtained some form of construction related education and training qualification. It can thus be deduced from the above analysis that the respondents’ representations are from SME owners who are mostly qualified and competent, and whose judgements and information provided towards the questionnaire survey can be considered reliable and valid.

4.2. Operational Years of SMEs

Results presented in

Table 8. Operational years of SMEs (source: [27,58]).

Variable		Frequency	Percentage	Cumulative Percentage
Number of years the SME owner’s organisation has been operational:	1–5 yrs	57	51.8	51.8
	6–10 yrs	27	24.5	76.4
	11–15 yrs	15	13.6	90.0
	16–20 yrs	3	2.7	92.7
	21–25 yrs	3	2.7	95.5
	More than 25 yrs	5	4.5	100.0
	Total	110	100

indicate that approximately 52% of the SMEs have been operational between 1 and 5 years, 25% between 6 and 10 years, and 14% between 11 and 15 years. Only 3% have been operational between 16 and 20 years, 3% between 21 and 25 years, and 5% have been operational for more than 25 years.

4.3. SMEs’ Cidb Class of Work and Cidb Grade

Table 9. SMEs’ cidb class of work and cidb grade (source: [27,58]).

Variable		Frequency	Percentage	Cumulative Percentage
cidb class of work that the SME organisation is registered in:	Both	50	45.5	45.5
	CE	31	28.2	73.6
	GB	29	26.4	100.0
	Total	110	100
cidb grade that the SME organisation is registered in:	Grade 1	46	41.8	41.8
	Grade 2	26	23.6	65.5
	Grade 3	10	9.1	74.5
	Grade 4	13	11.8	86.4
	Grade 5	6	5.5	91.8
	Grade 6	9	8.2	100.0
	Total	110	100

indicates that nearly 45% of the SMEs are registered for both the GB and CE classes of work: 28% in the CE class of work and 26% of SMEs registered as GB contractors. Table 5 also reveals that roughly 42% of the SMEs surveyed are grade 1 contractors and nearly 24% are grade 2 contractors; 9% of the SMEs stipulated that they are registered as grade 3 contractors, with approximately 12% indicating that they are registered as grade 4s, leaving almost 6% of the SMEs registered as grade 5 contractors and 8% as grade 6 contractors. Grades 1 to 6 are considered to represent the SME cluster as described by [82].

4.4. Measurement Model Results

The structural model with path coefficients and coefficient of determination R² values indicates that all variables selected for the development of the model were above the baseline of 0.5; therefore, no variables were deleted from the model. Based on the results reported in

Table 10. Latent variables’ inter-construct correlations and reliability measures (source: [58]).

Latent Variable	AVE	Composite Reliability	R2	Cronbach’s Alpha	Communality	Redundancy	AC	CC	CD	SP
CSR activities considered by SMEs to achieve SBP (ACD)	0.651	0.963	0.148	0.958	0.651	0.039	1	0	0	0
CSR implementation challenges experienced by SMEs (CC)	0.483	0.882	0	0.850	0.483	0	0.259	1	0	0
CSR drivers influencing the CSR practices of SMEs (CD)	0.780	0.946	0	0.931	0.780	0	0.245	0.139	1	0
SMEs’ perceptions pertaining to the relationship between the integration of CSR and SBP (SP)	0.745	0.959	0.101	0.952	0.745	0.035	0.287	0.239	0.241	1

, individual item reliability was inspected on the latent variables. The results show that the measures are robust in terms of their internal consistency reliability as indicated by the composite reliability. The composite reliabilities of the different measures ranged from 0.88 to 0.96, which exceeds the recommended minimum value of 0.70 stated by [93] cited by [94]. These results, according to Fornell et al. [95], confirm that convergent validity of the constructs may be concluded as adequate. The elements in the matrix diagonals, representing the square roots of the average variance extracted (AVE), are in all cases greater than the off-diagonal elements in their corresponding row and column, supporting the discriminant validity of the scales of use. The results illustrate higher factor loadings and the constructs indicate satisfactory shared variance with their indicators. Based on these observations, the model presents acceptable reliability and validity in explaining the links among the constructs of the model.

4.5. Validation of the Structural Model Results

Convergent validity was tested and confirmed by linking the latent variables in the model to extract the factor and cross-loadings of all indicator items to their respective latent variables. The structural model with path coefficients and coefficient of determination R² values (Figure 2) indicate that all items’ loads on their respective latent variables were from 0.63 to 0.91. In the proposed structural model (see Figure 2), interaction effects were examined by running the PLS algorithm to identify the relationship (if any) existing among the variables. The reason for running the PLS algorithm was to identify the variance explained by the variables included in the model and to establish the significance levels of the paths leading to the PLS estimate. The path coefficients were also evaluated and indicate the contributions of each latent explanatory construct to the predictive capacity of the endogenous construct. It is clear from Figure 2 that the exogenous and endogenous constructs of the model have a positive contribution towards each other. The overall predictive capacity of the structural model, according to Chin [89], is assessed by the R² value associated with the endogenous constructs within the model. Viewing the calculated R² value of the endogenous constructs, the values are above 10% which is acceptable, according to [96]. In addition, Frost [97] confirms that R² values vary in terms of the study area undertaken. Ref. [97] contends

Different research questions have different amounts of variability that are inherently unexplainable. Case in point, humans are hard to predict. Any study that attempts to predict human behaviour will tend to have R-squared values less than 50%. However, if you analyse a physical process and have very good measurements, you might expect R-squared values over 90%. There is no one—size fits all best answer for how high R-squared should be.

The R² values of this study aligned with the statement by [97] as the findings of the study were directly linked to the views of construction SME business owners in the SACI, allowing the R² to be assessed from a social science perspective.

To establish the significance level of the variables in this study, the bootstrapping technique part of the SmartPLS 2.0.M3 software was performed using 500 resamples. The bootstrapping technique produced Figure 3 which illustrates the structural model with path coefficients and t-statistics.

4.6. Structural Equations to Validate the Structural Model

The structural model (Figure 2 and Figure 3) gives an indication of how the latent variables link with each other. According to Monecke et al. [98] and Sanchez [99], latent variables specific to SEM can be segmented into two categories: endogenous and exogenous variables. Endogenous variables are influenced by one or more of the variables which form part of the model. Alternatively, exogenous variables are not influenced by other variables in the SEM. Based on SEM, endogenous variables portray dependency whereas exogenous variables portray independency.

With regards to the structural model (Figure 2 and Figure 3), CSR drivers influencing the CSR practices of SMEs and CSR implementation challenges experienced by SMEs are viewed as exogenous variables as these two variables are not influenced by variables in the model. It is important to note that the PLS-SEM structural model is considered a combination of linear regressions; thus, all the relationships in Figure 4 are linear, causal and additive [100]. As for the endogenous variables, the model (Figure 2) presents two endogenous variables (SME perceptions pertaining to the relationship between the integration of CSR and SBP, and CSR activities considered by SMEs to achieve SBP), with two sets of standardised coefficients estimated from the PLS-SEM. These PLS-SEM path equations relate to the causal link hypothesised in this study. The ε represents the error terms which denote that the variations remain unexplained by the predicting variables within the path model. The equations are as follows:

• $$\begin{array}{l}CSR implementation challenges experienced by SMEs \left(CC\right)= CC +\\ 0 \left(exogenous variable\right) \end{array}$$

  (4)
• $$\begin{array}{l}SME perceptions pertaining to the relationship between the integration of CSR and SBP \left(SP\right)=\\ PAB \left(CSR implementation challenges experienced by SMEs\right)+\\ PCB \left(CSR drivers influencing CSR practices of SMEs\right)+\\ \epsilon 1\end{array}$$

  (5)
• $$\begin{array}{l}CSR drivers influencing CSR practices of SMEs \left(CD\right)=CD +\\ 0 \left(exogenous variable\right)+\\ \epsilon 2\end{array}$$

  (6)
• $$\begin{array}{l}CSR activities considered by SMEs to achieve SBP \left(ACD\right)=\\ PAD \left(CSR implementation challenges experienced by SMEs\right)+\\ PBD \left(SMEs’ perceptions pertaining to the relationship between the integration of CSR and SBP\right)+\\ PCD \left(CSR drivers influencing CSR practices of SMEs\right)+\\ \epsilon 3\end{array}$$

  (7)

The following abbreviations represent the path coefficients, as illustrated in Figure 4

• PAB: CSR implementation challenges experienced by SMEs → SME perceptions pertaining to the relationship between the integration of CSR and SBP;
• PCB: CSR drivers influencing the CSR practices of SMEs → SME perceptions pertaining to the relationship between the integration of CSR and SBP;
• PAD: CSR implementation challenges experienced by SMEs → CSR activities considered by SMEs to achieve SBP;
• PBD: SMEs perceptions pertaining to the relationship between the integration of CSR and SBP → CSR activities considered by SMEs to achieve SBP;
• PCD: CSR drivers influencing the CSR practices of SMEs → CSR activities considered by SMEs to achieve SBP.

Figure 5 illustrates the CSR model that has been established to guide SMEs in the SACI towards the achievement of SBP. The model constructs and measurement variables for each construct are shown in the model.

4.7. Model Evaluation

With regards to the evaluation of the model developed, the PLS considers the R-squared values as essential to determine the predictive significance of the model. Considering the ongoing need to report and evaluate the performance of PLS models, including both measurement and structural models, and with attention on the overall predictive power of the model, a global criterion of goodness of fit (GoF) index as recommended by [91] was used. The procedural guidelines provided by [92] to compute the GoF values, which are minimum values for global validation of PLS path models, were followed. Based on the values (Table 6), a GoF value of 0.29 was achieved for the entire model, falling within the threshold values of 0.25 and 0.36 for small and large values as stipulated by [101]. Based on this, it can be concluded that the PLS model developed in this study has explanatory power and offers support to validate the PLS model globally.

5. Discussion of Findings from the Model Results

The results from the structural model developed indicate that the CSR implementation challenges experienced by SMEs in the SACI, along with the CSR drivers influencing the CSR practices of SMEs, have a predictive power of 10.1% in terms of influencing SME perceptions pertaining to the relationship between the integration of CSR in their businesses and SBP. According to the report of [96], an R² value of 10.1% is considered acceptable. CSR implementation challenges experienced by SMEs explained 21% and CSR drivers influencing the CSR practices of SMEs explained 21.1% towards SME perceptions regarding the relationship between the integration of CSR in their businesses and SBP. Further analysis shows the following: a lack of integration in SME culture and SME business objectives and norms; limited financial resources to undertake CSR initiatives; limited human resources to undertake CSR initiatives; lack of CSR skills and knowledge; difficulty adapting CSR practices and standards to internal business processes; unstable economic conditions; and poor collaboration among SMEs. These indicators contribute to the significance of CSR implementation challenges experienced by SMEs on various management levels which influence the perceptions of SMEs regarding the relationship between the integration of CSR in their business and SBP, as summarised by the model (path (r) = 0.210; t = 1.805; p ≤ 0.10).

In terms of CSR drivers influencing the CSR practices of SMEs, it is evident that several drivers—global standardisation; stakeholder activism; socio-economic priorities and concerns; political reforms; and culture and tradition—contribute substantially as CSR drivers which in turn influence SME perceptions in the SACI based on the relationship between the integration of CSR and SBP, as summarised by the model (path (r) = 0.211; t = 2.453; p ≤ 0.05). The results pertaining to CSR implementation challenges and CSR drivers influencing CSR practices of SMEs are supported by The Peak Performance Centre [54] who argues that past and present challenges have a direct influence on the perceptions of people and impact on decision-making processes. According to Zhang et al. [20], this information is correlated to the way in which business owners, particularly SME construction business owners, perceive their business environment and the way business decisions around CSR initiatives and activities are made, taking into consideration CSR drivers and implementation challenges which influence their perceptions of CSR practice.

In addition, the model examines the relationships between SME perceptions: the relationship between the integration of CSR in their businesses and SBP, CSR drivers influencing CSR practices of SMEs, CSR implementation challenges experienced by SMEs and the CSR activities considered by SMEs to achieve SBP. The model indicates that SME perceptions pertaining to the relationship between the integration of CSR and SBP explained 20.1% towards the CSR activities considered by SMEs, which include among other things employee rights, remuneration and recruitment; permissible shareholder proceeds; customer satisfaction and product safety; preserving suitable supplier and partner relationships; and conformance to the requirements of government laws and policies. The model also indicates that SME perceptions have a positive significant relationship with CSR activities considered by SMEs to achieve SBP, as summarised by the model (path (r) = 0.201; t = 2.067; p ≤ 0.05). This is supported by Hurley [55] who references Gibson’s theory of perception wherein perception is viewed as a requisite property of animate action, arguing that without perception being realised, action (in this case the decision to undertake CSR activities) would be unguided and, without action, perception would serve no purpose. UK Essays [56] concur, arguing that decision making is an important skill that a business owner must exercise for the business to achieve business goals and objectives. Elford et al. [59] further mentions that organisational excellence, which includes SBP, leans heavily on proper decision making (in this case, the decision to undertake CSR activities) by the business owner and management team, guided by their perceptions.

The model also indicates that CSR implementation challenges experienced by SMEs explained 18.7% towards CSR activities considered by SMEs to achieve SBP. The CSR implementation challenges experienced by SMEs share a positive significant relationship with the CSR activities considered by SMEs to achieve SBP as summarised by the model (path (r) = 0.187; t = 1.684; p ≤ 0.10). Zhang et al. [20], The Peak Performance Centre [54], Elford et al. [59], and Loosemore and Loosemore [65,66] support the results. Lastly, the model indicates that CSR drivers influencing CSR practices of SMEs explained 17.1% toward CSR activities considered by SMEs to achieve SBP. CSR drivers influencing the CSR practices of SMEs share a positive significant relationship with the CSR activities considered by SMEs to achieve SBP as summarised by the model (path (r) = 0.171; t = 1.973; p ≤ 0.05). Studies by [20,69,81] support the results.

In summary, the reflected results based on the structural model illustrate that CSR implementation challenges experienced by SMEs, and CSR drivers influencing CSR practices of SMEs, have positive significant relationships and moderate predictive capabilities to influence SME perceptions pertaining to the relationship between the integration of CSR in their businesses and SBP. This is similar for the relationships and predictive capability which SME perceptions pertaining to the relationship between the integration of CSR in their businesses and SBP, CSR drivers influencing CSR practices of SMEs and CSR implementation challenges experienced by SMEs have on the CSR activities considered by SMEs to achieve SBP, amounting to 14.8%. The overall predictive strength of the CSR model is acceptable as the R² values are above 10%. The accepted predictive strength of the model has thus supported the research hypotheses stipulated in Table 2.

Table 11. Summary of the effects of structural model results on hypothesised links in PLS-SEM path model (source: [58]).

Path Label	Path Relationship	T-Statistic	Corresponding Hypothesised Path	Remark on Hypothesis
PAB	CSR implementation challenges experienced by SMEs → SMEs’ perceptions pertaining to the relationship between the integration of CSR and sustainable business performance	Significant	Hypothesis 1: there is a significant relationship between CSR implementation challenges experienced by SMEs and SMEs’ perceptions pertaining to the relationship between the integration of CSR and sustainable business performance	Supported
PCB	CSR drivers influencing the CSR practices of SMEs → SMEs perceptions pertaining to the relationship between the integration of CSR and sustainable business performance	Significant	Hypothesis 2: a significant association exists between CSR drivers influencing the CSR practices of SMEs and SMEs’ perceptions pertaining to the relationship between the integration of CSR and sustainable business performance	Supported
PAD	CSR implementation challenges experienced by SMEs → CSR activities considered by SMEs to achieve sustainable business performance	Significant	Hypothesis 3: there is a significant relationship between CSR implementation challenges experienced by SMEs and the CSR activities considered by SMEs to achieve sustainable business performance	Supported
PBD	SMEs’ perceptions pertaining to the relationship between the integration of CSR and SBP→ CSR activities considered by SMEs to achieve sustainable business performance	Significant	Hypothesis 4: a significant affiliation exists between SMEs’ perceptions pertaining to the relationship between the integration of CSR and SBP and the CSR activities considered by SMEs to achieve sustainable business performance	Supported
PCD	CSR drivers influencing the CSR practices of SMEs → CSR activities considered by SMEs to achieve sustainable business performance	Significant	Hypothesis 5: there is a significant relationship between CSR drivers influencing the CSR practices of SMEs and the CSR activities considered by SMEs to achieve sustainable business performance	Supported
PAD+PBD+PCD	Combined paths	Significant	Hypothesis 6: merging the CSR implementation challenges experienced by SMEs, and SMEs’ perceptions pertaining to the relationship between the integration of CSR and sustainable business performance, as well as CSR drivers influencing the CSR practices of SMEs, impacts the CSR activities considered by SMEs to achieve sustainable business performance	Supported

summarises the effects of the structural model results on the hypothesised links in the PLS-SEM path model.

6. Conclusions

This study acknowledges that the understanding of what CSR means to the construction industry, and how it is practiced, is still limited as little research has been undertaken to develop a framework for CSR activities relevant to construction enterprises worldwide as a tool for CSR performance and ultimately SBP for construction enterprises large or small. This limitation is supported by [19] and [20] cited by [27]. Moreover, a limitation pertaining to a CSR model to guide SMEs, particularly in the SACI, towards SBP has subsequently also been identified, considering the research conducted by [19,20,22,24,27]. On this premise, the contribution of this study was establishing the following:

• That SMEs in the SACI perceive a positive relationship between the integration of CSR within their business and sustainable business performance;
• That, although limited, CSR practices of SMEs in the South African construction industry are driven by certain international and national CSR drivers;
• That SMEs in the SACI face CSR implementation challenges across all management levels pertaining to the organisation and the business environment;
• That SMEs in the SACI consider specific CSR activities across nine CSR dimensions (employees; shareholders; customers; suppliers and partners; government; environment and resources; community; competitors; and NGOs) to achieve sustainable business performance.

A further and major contribution was the development of a novel CSR model to guide SMEs in the SACI towards achieving sustainable business performance, utilising a ‘Partial Least Squares Structural Equation Model’. The model was validated through hypothesis testing. The suitability of PLS-SEM was attested by [102], that PLS-SEM is a strong method for research that intends to refine theories in management research because it offers a variety of advantages. Thus far, though, limited use of PLS-SEM has been observed in construction management research, more specifically in research relative to the concept of CSR in the global construction industry. However, this study has illustrated that PLS-SEM is a crucial multivariate method of analysis that can advance the study of CSR and sustainable business performance in modelling relationships of variables. The model therefore indicates the following:

• That CSR implementation challenges experienced by SMEs across all management levels pertaining to the organisation and business environment significantly influence the perception of SMEs relative to the relationship between the integration of CSR and sustainable business performance, which in turn significantly influences the CSR activities considered by SMEs to achieve sustainable business performance;
• That international and national CSR drivers influencing the CSR practice of SMEs significantly influence the perception of SMEs relative to the relationship between the integration of CSR and sustainable business performance, which significantly influences the CSR activities considered by SMEs to achieve sustainable business performance;
• That, individually, CSR implementation challenges experienced by SMEs across all management levels pertaining to the organisation and business environment; SME perceptions relative to the relationship between the integration of CSR and sustainable business performance; and international and national CSR drivers influencing the CSR practice of SMEs all significantly influence the CSR activities considered by SMEs to achieve sustainable business performance.

The developed CSR model combined two theories, namely perception theory and stakeholder theory, to support the CSR model. This is novel as other CSR research has overlooked perception theory as a catalyst to stakeholder theory. Based on this theoretical implication it should be noted that the CSR model developed is intended for practical use and therefore recommendations are directed towards government agencies such as the cidb, policy makers and CETA as well as institutions of higher learning which are housed in the South African context. Hence the following recommendations are made: policy makers in government should assist by phasing in more enforceable statutory requirements in line with the adoption of CSR, that will be utilised as a guide for training and monitoring mechanisms, ensuring the achievement of SBP of SMEs in the SACI; to guide SMEs in the SACI towards achieving SBP from a CSR perspective, it is important for CETA and institutions of higher learning to assist government by developing and administering accredited CSR training programmes for construction SMEs, and, as a government agency, the cidb should assist by introducing a CSR merit and demerit monitoring system for the development of SMEs in the SACI, ultimately driving SMEs to perform CSR activities that are proven by this study to contribute to the achievement of SBP.

Further practical limitations were encountered relative to the questionnaire survey and structured interviews utilised for this study. Both considered only SMEs in the SACI who are registered on the cidb register of contractors between Grade 1 GB or CE and Grade 6 GB or CE which means the results may only be valid for the South African context, though the generic methodology, data analysis techniques and the model can be replicated for other countries. Another limitation was the difficulty in collecting data during the COVID-19 pandemic. This manifested itself in that many SMEs in the SACI prioritised their business survival over research participation, which is well understood. This resulted in time, administrative and financial constraints experienced during the research. Despite this, sufficient data was obtained to validate the findings, particularly the developed CSR model.

It is also proposed that further research be conducted on the following topics: a thorough investigation into why SMEs in the SACI are limitedly driven by CSR drivers; modalities that can be utilised in mitigating CSR implementation challenges identified in this research; whether relationships exist between the CSR drivers influencing the CSR practice of SMEs and the CSR implementation challenges experienced by SMEs in the SACI, and vice versa; identifying and establishing an appropriate CSR module that could be embedded in training programmes aimed at developing SMEs in the SACI; the development of concise statutory requirements and monitoring systems for the practice of CSR within SMEs and larger construction organisations within the SACI; the constructs of the CSR model developed in this study and their corresponding variables could be explored as theoretical views and various related topics could be considered developmental.