Circular procurement implementation barriers in the construction industry of developing economies: a quantitative analysis Open Access

The construction industry significantly enables economic development and citizens’ well-being globally (Otasowie et al., 2023; Ikuabe et al., 2023a, b). The industry achieves this feat by procuring and delivering critical infrastructure and providing employment to a significant proportion of the populace, whether unskilled, semi-skilled or skilled. According to the World Economic Forum (2024), the construction industry contributes about 13% to the global GDP, whilst being responsible for the direct and indirect employment of a significant proportion of the worldwide population. Despite these critical contributions towards advancing society’s economic and social growth aspirations, the negative impact of a plethora of anthropogenic activities inherent in the industry’s project delivery ecosystem on attaining sustainable futures remains a growing concern. The proclivity of these activities towards the linear economy model of production and consumption, otherwise referred to as the take-make-dispose model, has been blamed for the industry’s unsustainable utilisation of natural raw materials and disposal of construction and demolition waste (Adekunle et al., 2023; Ikuabe et al., 2024; Ramakrishna et al., 2020). Due to the preponderance of these anthropogenic activities within the construction projects, the construction industry is responsible for using three billion tonnes of raw materials whilst generating waste in like sum (Ramakrishna et al., 2020). Also, the industry is known for emitting significant amounts of carbon and greenhouse gases, further undermining society’s carbon neutrality ambitions and exacerbating global warming (Stephen et al., 2025). Admittedly, these negative impacts are exacerbated by the growing population and rapidly urbanising cities, which have heightened the demand for buildings and infrastructure, which would, in turn, culminate in the increased use of the earth’s resources and greenhouse gas emissions. This reality has necessitated the adoption of various interventions by relevant stakeholders within the context of the construction industry to ameliorate these challenges. Adopting circular economy (CE) principles across the construction project lifecycle has been identified as a veritable pathway for achieving sustainable performance in such projects. Although defying a widely accepted definition, the CE concept remains synonymous with processes associated with eliminating waste (Ogunmakinde et al., 2022). The CE concept relies on the 10 Rs principle to minimise the reliance on virgin materials during production. It provides a closed-loop system wherein materials and assets are used for longer intervals and, upon decommissioning, deployed to alternative uses as inputs into another production system (Ramakrishna et al., 2020). The growing population and associated rapidly urbanising cities-construction industry has adopted several initiatives to improve its sustainability performance. This gives the premise for circular procurement (CP) for construction project delivery. CP is a strategic approach to purchasing that supports the transition to a CE (Neessen et al., 2021). It is a system that prioritizes key concepts such as repair, reuse, recycling, refurbishment and sustainability, instead of employing the traditional linear procurement model. Based on this background, this study is motivated to explore the barriers to implementing CP in the construction industry from a developing economy perspective.

Highlighted in existing literature, CE has been identified as contributing to construction sustainability through significant changes in the industry’s supply chains and procurement processes (EMF, 2017). In literature, the integration of CE and procurement is known as CP (Sönnichsen and Clement, 2020), which holds immense potential for the construction industry. Construction CP is regarded as the procurement strategy that public and private organisations adopt to pursue CE actively (Neessen et al., 2021). The United Nations Environment Programme (UNEP) defines “circular procurement as when the buyer purchases products or services that follow CE principles, supporting the assessment of designing, making, selling, reusing and recycling products to determine how to get the maximum value from them, both in use and at the end of their life” (UNEP, 2021). The current economic production and consumption model satisfies procurement activities’ needs while compromising future needs. Geissdoerfer et al. (2017) argue that this economic model results in enormous quantities of rare resources ending up in landfills, as most materials are discarded after the product’s lifetime. In addition, procurement teams typically favour using new rather than recycled resources to meet the requirements of this conventional linear procurement system. An alternative to this model is CP, which aims to keep materials in the supply chain longer (Kirchherr et al., 2017). Thus, materials are kept in circulation for an extended period for continuous value extraction. CP aims towards a zero-wastage concept even after the product’s end of life through forward and reverse supply chains (Antikainen and Valkokari, 2016) and further suggests cross-sector collaboration to maximise the value of products and services (Qazi and Appolloni, 2022). CP also significantly contributes to the industry’s compliance with some international regulations such as Sustainable Development Goals 11 and 12 – (“sustainable cities and communities” and “responsible consumption and production”). This is achieved through the development of sustainable urban infrastructure by prioritizing materials and systems that are reusable, durable and resource-efficient. In addition, it directly fosters responsible consumption and production by embedding life cycle thinking into purchasing decisions.

In addition, implementing CP practices can be complex as it involves tangible and intangible process activities for transforming traditional methods. However, three levels of adoption of CP have been found in the research studies. First and foremost, at the “system level”, the focus is on encouraging contractual servitisation rather than owning. It stresses the benefits of renting or leasing methods and collaborating with other organisations to share. Secondly, “the supplier level” explains suppliers’ roles in integrating circularity into resources and processes. The system supports the design for disassembly, supplier trackback mechanism and other features. The final level pertains to the product and the re-utilisation of materials or components after their initial use (European Union, 2017). Construction CP encompasses more than just providing construction products and services. It considers the broader and intricate network of stakeholders and supply chain activities, considering the entire lifespan of a project (Ababio and Lu, 2023). Tender design, a vital part of the construction process, is also crucial in CP implementation. CP incorporates circularity and environmental effectiveness principles into client specifications throughout the tendering process, post-award contract administration, and organisational and government procurement policies (Alhola et al., 2019). Implementing CP within industrial processes holds great potential by fostering extensive and continuous stakeholder engagements, creating new business models and allocating efficient roles to improve the regeneration and restoration of material stock within systems (Marrucci et al., 2019). Again, CP relies on desirable components of conventional procurement processes, such as value-for-money prioritisation, quality concerns and environmental consequences, to create a circulation system of energy and materials flow in a closed loop while staying within sustainability boundaries (van Oppen et al., 2018). Over the past few years, the use of CP in public procurement for construction has increased significantly, with developed countries accounting for a larger share of its implementation than developing countries (Kristensen et al., 2021). Table 1 summarises a list of barriers to adopting CP within the construction industry.

The study assessed the inhibitors to implementing CP by construction organisations in Nigeria. This was actualised using a deductive approach hinged on a post-positivism philosophical viewpoint aided by quantitative data elicited from construction professionals using a questionnaire survey. The use of a questionnaire is due to its ability to elicit responses from a large pool of respondents and allow the quantifiability and objectiveness of the study (Tan, 2011). The research area of the study was Lagos and Abuja, Nigeria. The choice of both locations stems from their commercial and administrative attributes. Lagos serves as the commercial capital, while Abuja is the country’s seat of power, with both cities boasting many construction professionals and attributed to numerous construction projects of varying sizes. The target respondents of the study were construction professionals, namely builders, architects, quantity surveyors, engineers and project managers. The sample size employed for the study was derived using the formula given by Yamane (1967) – n = N/1 + N (e) $2$

The data was used after the derivation of the number of registered professionals, which made up the sample frame of the study. Thereafter, one hundred and seventy-four questionnaires were distributed, while one hundred and sixty-three were returned, and all were filled out correctly. The sampling technique used in the study was the cluster probability sampling technique. The chosen technique was achieved by getting hold of the various registered professionals of the professional bodies in Lagos and Abuja, Nigeria. The researcher distributed the questionnaire electronically, which expedited the process of data acquisition from the target respondents. Before collecting data from the target respondents, a pilot study was conducted, which helped frame the needed components for the research instrument. Also, the validity and reliability of the research instrument were ascertained using Cronbach’s alpha test. The test gave an alpha value of 0.912, which affirms the validity and reliability of the survey instrument (Tavakol and Dennick, 2011).

The review of extant literature identified twenty-three inhibiting factors to implementing CP by construction organisations in Nigeria. The data retrieved from the respondents of the study were analysed using the following methods of data analysis: mean item score (MIS), Kruskal-Wallis h-test (K-W), Student Newman-Keuls (SNK) post hoc technique, exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). The MIS was used to rank the identified inhibitors based on the opinions of the study’s respondents. Kruskal-Wallis h-test (K-W) was employed to ascertain if there is a statistical difference in the views presented by the respondents based on their professional affiliation. K-W is a non-parametric test used in evaluating the significant difference in the opinions of more than two groups (Pallant, 2005). The outcome of the K-W test provides the chi-square and p-value. When the given p-value >0.05, it is deemed that there is no significant difference in the opinions of the group of respondents. In contrast, when the given p-value <0.05, it is deemed that there is a significant difference in the views of the group of respondents. In addition, the SNK post hoc test was deployed to differentiate the mean responses of the respondents’ classification based on their professional designation. Furthermore, the study used principal component analysis (PCA) as the extraction method in the conduct of EFA. This aided in determining the unidimensionality and factor analysisability of the identified inhibitors, as suggested by Oke et al. (2021) and Ikuabe et al. (2022). The idea of using EFA is driven by providing insights and clarity into structured patterns, giving a more detailed understanding of the connection patterns of the variables (Young and Pearce, 2013). Using the EQation software (EQS) version 6.4, the obtained constructs from the EFA were subjected to further analysis to affirm their equivalency. This was executed with the aid of CFA. The study used a multi-dimensional technique to assess the model. This was situated within the purview of the following fit indices: Satorra–Bentler scaled Chi-square (S – Bχ²), root mean square error of approximation (RMSEA), standardised root mean square residual (SRMR) and root mean square error of approximation with 95% or 90% confidence interval (RMSEA @ 95% or 90% CI), goodness-of-fit index (GFI) and Bentler comparative fit index (CFI). According to Kaplan (2009), these indices outline a comparative absolute fit index and incremental fit indices. Furthermore, these indices exhibit a thorough and detailed assessment of the suitability of the derived constructs with the data sample. Moreover, further evaluation conducted by the analysis includes the z-statistics, construct validity and internal consistency.

The information on the respondents of the study shows that architects and quantity surveyors made 33.7% and 27.9% of the respondents, respectively. In contrast, builders and engineers comprised of 15.5% and 12% of the respondents, respectively. Also, 10.9% of the respondents were project managers. In terms of the years of professional experience of the respondents, 27.4% attained 11–15 years of experience, 19.4% attained 6–10 years of experience, 18.7% attained 1–5 years of experience and 22.8% of the respondents have above 15 years of professional experience, while 11.7% have less than 1 year of professional experience.

The review of related literature highlighted twenty-three inhibiting factors to implementing CP. These were presented to the target respondents of the study for rating based on their significance. The result of the analysis conducted on the retrieved data is presented in Table 2. The results show the ranking of the inhibiting factors and the K-W test. The result indicates that the most ranked factor is no commitment from top management, with a mean score of 4.68. The next ranked factors are lack of financial resources and perceived high cost associated with CP, with mean scores of 4.55 and 4.51, respectively. This is closely followed by divergent procurement priorities and lack of legislation, with mean scores of 4.43 and 4.40, respectively. The least ranked factors are non-conformance with environmental laws and lack of return pathway, with mean scores of 3.38 and 3.44, respectively. All the inhibiting factors are observed above 3.00, which portrays their significance. Furthermore, the outcome of the K-W test showed no significant difference in the opinions of the respondents based on their professional affiliation with nineteen factors. These nineteen factors have a p-value above 0.05, which implies a convergence in the views of the group of professionals. Conversely, there is a divergent stance in the opinions of the group of professionals on four factors. These factors are divergent procurement priorities, unclear strategic goals among project team members, inadequate training, and non-involvement of suppliers at an early stage. These factors have a p-value of less than 0.05.

Table 3 outlines the result of the SNK post-hoc test conducted in the study. The outcome of the K-W test indicates that there are divergent opinions on four factors serving as inhibitors to the implementation of CP by construction organisations in Nigeria. Hence, it becomes imperative to provide details about the different viewpoints. The SNK post hoc test (multiple comparisons) outlines the mean difference between the various groups of respondents; in this case, it is a professional designation. The findings show a difference in opinion among the three classifications of the professional designation. The first group comprises builders and engineers with values of 2.8117 and 2.6834, respectively. The second group includes quantity surveyors and project managers with values of 2.7295 and 2.5052, respectively. The third group is made of architects with a value of 2.5263.

EFA was employed to cluster the identified factors inhibiting the implementation of CP by construction organisations into more manageable constructs. This was achieved using PCA with varimax rotation. According to Tabachnick and Fidell (2007), this extraction technique aids in reducing large numbers of variables into coherent components or clusters. Hair et al. (2006) affirmed that EFA’s conduct requires a large sample size for reasonable results; therefore, the study used one hundred and sixty-three as the sample size. To ascertain the factorability of the dataset, the Kaiser–Meyer–Olkin (KMO) of sampling adequacy and Bartlett’s test of sphericity were employed. The result of the KMO sampling adequacy and Bartlett’s test of sphericity is presented in Table 4. It is revealed that the KMO value given is 0.965, above the threshold of 0.6 used by previous studies (Aghimien et al., 2021; Ikuabe et al., 2021). Also, Bartlett’s test of sphericity gave a value of 1137.142 and a p-value of 0.000, thus implying that it is significant (Pallant, 2005). The results affirm the suitability and factorability of the study’s dataset for EFA.

Table 4 also presents the results of the factor loadings emanating from the EFA. It is revealed that the identified twenty-three factors are loaded into five components, which are achieved through PCA and the adoption of varimax rotation as the extraction technique. The rotated component matrix’s outcome shows that the factor loading values are all above the 50% threshold as adopted by previous studies (Aghimien et al., 2021; Ikuabe et al., 2022). These results, in combination with the extracted communalities, show a good correlation between the formed components yielded by the analysis. Five components from the analysis are extracted, as shown in the table. The first component exhibits five variables whose factor loadings range from 0.927 to 0.635 and are named inadequacy of government policies and initiatives, while the second component has six variables with factor loadings that range from 0.851 to 0.551 and are named stakeholder-related challenges. The third component is attributed to five variables having factor loadings from a range of 0.728 to 0.524, named procurement bottlenecks. In comparison, the fourth component has four variables with a factor loading ranging from 0.708 to 0.517 and is named organisational culture. The fifth component comprises three variables with factor loadings ranging from 0.672 to 0.506 and is named financial impediments. The derived components’ names emanate from the intrinsic attributes of the variables that formed the various components. Also, the percentage of the variance explained by the individual components is presented in Table 4. It is revealed that components one, two and three have 34.18%, 17.22% and 9.92, respectively. Meanwhile, components four and five have 5.06% and 3.84% respectively. The cumulative sum of the percentage variance explained for all components is 70.22%.

The findings of the EFA were further subjected to assessing the validity of the resulting components using CFA. The study’s dataset is attributed to non-normality. Hence, the study employed the robust maximum likelihood estimation technique. This estimation method caters to the non-normality of the dataset (Field, 2009). The standardised coefficient ((λ), the z-statistics, the coefficient of determination (R²) and the validity of the derived components are presented in Table 5. Also, the Rho alpha and Cronbach alpha were employed to assess the model’s internal consistency. Cepeda-Carrion et al. (2019) noted that using both tests is advised in determining the model reliability when evaluating the internal consistency for attaining a robust output. The result shows that the standardised coefficient of the variables ranges from 0.927 to 0.634. This portrays a good construct validity as all the variables are well above 0.5 (50%) of the variance, which is the threshold. Furthermore, it is indicated that Cronbach's alpha and Rho’s alpha tests gave a resulting coefficient of 0.881 and 0.874, respectively. Hair et al. (2019) noted that it is recommended for the coefficients to be above 0.7 and closer to 1.00, hence indicating good reliability from the derived constructs. The outcome of the z-statistics shows that all the variables have a value greater than 1.96. This reinforces the significance of the identified variables as inhibiting factors to implementing CE construction organisations. Moreover, the results of the R² show that the values range from 0.871 to 0.591. This affirms the potent predictive accuracy of the variables. Henseler et al. (2009) noted that R² values closer to 1.0 indicate a substantial predictive accuracy. Also, the group R² of the derived constructs shows that the most potent predictive accuracy constructs are procurement bottlenecks, the inadequacy of government policies and initiatives, and financial impediments with values of 0.826, 0.793 and 0.784, respectively. This is followed by stakeholder-related challenges and organisational culture with values of 0.771 and 0.707, respectively.

The result of the fit indices and the derivative output estimate is presented in Table 6. It is shown that the CFI and GFI values derived are 0.957 and 0.968, respectively. This informs that both indices attain a good fit, as it is noted that a good fit is achieved when the resulting value is ≥ 0.95, while an acceptable fit is achieved when the resulting value is ≥ 0.90 (Iacobucci, 2010). Furthermore, it is shown that the resulting values of the RMSEA and SRMR are 0.045 and 0.072, respectively. This outcome indicates that the RMSEA attained a good fit while the SRMR attained an acceptable fit. Bentler (2005) recommended a good fit for both indices when the resulting value ≤ 0.05, while an acceptable fit is recommended when the resulting value ≤ 0.08. Also, the resulting analysis value of the model’s sample data is an S – Bχ² of 5.431 with 2° freedom and an associated p-value of 0.000. Due to the high sensitivity associated with chi-square in sample size and the data’s normality, it is adjudged less reliable (Zhong and Yuan, 2011). Consequently, it is encouraged that the normed chi-square is employed (Kline, 2005). This is achieved by using the degree of freedom to divide the chi-square. Hence, the normed chi-square from the study is 2.716. Byrne (2006) recommended values ranging from 3.00 to 5.00 for a good fit. Therefore, the resulting normed chi-square of this study affirms the model’s adequacy.

The analysis indicates that the most significant factors inhibiting the deployment of CP are no commitment from top management, lack of financial resources, perceived high cost associated with Circular Products and Services, and divergent procurement priorities. Also, the result of the CFA conducted using the derived value of R² indicates that the inadequacy of government policies and initiatives is the most significant construct that outlines the implementation barriers of CP.

The study’s analysis shows that the inadequacy of government policies and initiatives significantly hinders the implementation of CP. This component comprises lack of legislation, government support, non-conformance with environmental laws, conflicting existing policies and lack of political will for implementation. This finding is in tandem with Tura et al. (2019) and Munaro and Tavares (2023), who noted that the lack of circular-specific laws or regulations hinders the implementation of CP strategies in the construction industry. Also, it has been reported that existing policies create barriers that impede the harmonisation and operationalisation of CP practices (Kristensen et al., 2021; Lahane and Kant, 2021). The lack of government initiatives significantly hampers the drive to implement CP. Without clear policies directing initiatives such as using recycled materials, resource efficiency and waste reduction, the concept of CP would be a mirage in the construction industry. Moreover, the lack of standardised guidelines further generates uncertainty (Kumar Mangla et al., 2021; Munaro and Tavares, 2023), which leads to the hurdle of general acceptance. Furthermore, insufficient political will impedes the commitment to overcoming CP implementation barriers. This is corroborated by Tura et al. (2019), who noted that effective policies backed by firm enforcement are important in driving the implementation of CP in the construction industry.

The second component derived from the analysis is labelled stakeholders-related challenges. It comprises the following variables: lack of awareness, inadequate training, unwillingness to implement CP, incapacity of suppliers, unclear strategic goal among project team members and unclear integration plan among supply chain partners. This finding aligns with the study of Ahmed et al. (2024), which affirmed that a deficiency in organisational and industrial awareness of CP hinders its adoption and implementation. The absence of clear communication and shared objectives hinders the drive for CP in construction project delivery. Furthermore, a deficiency in organisational and stakeholders’ awareness of CP inhibits its acceptance and implementation (Ababio and Lu, 2023; Munaro and Tavares, 2023). It becomes imperative to outline the need for intensive sensitisation of CP awareness among stakeholders in the construction industry. In addition, weak collaboration between stakeholders results in disjointed decision-making, which leads to ineffectiveness in material recycling and reuse. Moreover, the lack of clarity on strategic planning for integrating supply chain partners hinders CP implementation within the industry (van Keulen and Kirchherr, 2021).

The third component, which comprises five variables, is called procurement bottlenecks. The variables are divergent procurement priorities, non-involvement of suppliers at an early stage, lack of return pathway, knowledge of return mechanism and sourcing not intended for the return of material. Ahmed et al. (2024) noted that differences in procurement objectives or preferences of involved stakeholders obstruct CP implementation. Usually, traditional processes prioritise cost-effectiveness and speedy completion over long-term value and sustainability. Moreover, the late involvement of suppliers in the procurement process tends to hinder service and product selection criteria for the attainment of CP (De Angelis et al., 2018). Furthermore, supply chain limitations outline hurdles associated with a lack of supply of circular materials. Qazi and Appolloni (2022) affirm that the complexity of the reverse supply chain poses challenges to procurement processes, encouraging firms to source non-returnable materials. Many suppliers lack the capacity or incentives to produce and distribute recycled or repurposed construction materials at competitive prices. This is further exacerbated by lengthy approval processes, which serve as a significant challenge for procurement decisions.

The fourth component, labelled as organisational culture, comprises four variables. The variables are no commitment from top management, bureaucratic organisational system, unclear pathway to sustainable construction integration and culture of risk avoidance. This finding is corroborated by Leal Filho et al. (2019), who noted that organisational systems prioritising cost and quality metrics over environmental and social considerations present obstacles to CP implementation. Many construction organisations have a defined operational environment that revolves around cost-driven tenets that place short-term financial gains over attaining long-term sustainable principles. Moreover, the lack of commitment from organisational leadership in steering the drive towards CP is a significant obstacle (Farooque et al., 2019; Qazi and Appolloni, 2022; Mangla et al., 2018; Haselsteiner et al., 2021). When top management in organisations does not actively promote CE principles for procurement activities, there is a great tendency for the organisation not to imbibe the sustainability concepts.

The last component derived from the analysis comprises three variables labelled as financial impediments. The variables are lack of financial resources, high costs associated with returned materials and perceived high costs related to circular products and services. This finding is in tandem with the study of Farooque et al. (2019), who stated that the insufficiency of an organisation’s financial means and resources to support CP practices impedes its implementation. Deploying sustainable materials and recycling processes requires substantial initial investment, which might be difficult for many construction organisations to key into. This becomes more challenging in severely competitive markets where cost optimisation is a significant priority. Also, the cost of processing returned materials to meet quality standards for reuse or recycling is a barrier to organisations’ widespread adoption of CP (Lahane and Kant, 2021).

The study assessed the factors inhibiting the implementation of CP from a developing economy perspective. Using a quantitative approach, data were elicited from construction professionals and analysed using multiple analysis techniques. The findings showed that the most significant barriers to implementing CP are no commitment from top management, lack of financial resources, and perceived high costs associated with circular products and services. Also, the EFA findings revealed five constructs: inadequacy of government policies and initiatives, stakeholder-related challenges, procurement bottlenecks, organisational culture and financial impediments. The CFA conducted for the study further affirmed these derived constructs using construct validity, reliability and fit indices. With the study’s outcome, it is essential to state that the drive for implementing CP can be achieved using a multi-faceted approach. The government must commit to establishing regulatory frameworks and policies that clearly support and mandate the deployment of CP. Also, stakeholder engagement is paramount in overcoming resistance to change. This can be actualised through awareness campaigns and constant training programs that help educate and train various stakeholders associated with the delivery of construction projects.

The study’s outcome contributes immensely to the growing conversation on sustainable development. Theoretically, it contributes to the body of knowledge by filling the gap in scholarly studies centred on sustainable procurement from a developing economy perspective. This will serve as a solid theoretical foundation for further studies geared towards contributing to the dialogue on sustainable development by implementing viable procurement concepts and techniques. Moreover, from a practical perspective, the study offers valuable insights that can help abate the critical barriers to the implementation of CP in construction project delivery. The outcome of the study can assist relevant stakeholders involved in the procurement of construction projects in making informed decisions that can contribute towards sustainable development from a developing economy context.

It is important to state that the study was conducted in only two cities in Nigeria, Lagos and Abuja. Future studies can be conducted in other cities of the country to get a broader perspective on the barriers to CP implementation. Also, the current study employed a quantitative approach; future studies can employ a qualitative method or mixed methods, as this might bring a different perspective to the assessment of implementation barriers of CP for construction project delivery.