Project policy and project outcomes: The moderating role of project leadership

Abstract Ghana continues to record increasing numbers of uncompleted projects with its attendant cost overruns which are borne by the taxpayer. Several studies have sought to explain this phenomenon by exploring how project policy influences project outcomes. Though project leadership have been established to contribute to project outcomes, a study to look at how project leadership interplays between project policy and project outcome are lacking, hence the focus of this study. The current study thus uses PLS-SEM to model how project leadership interacts with project policy and project outcomes using government projects within the Cape Coast Metropolitan area. The study uses the resource-based view theory to explain the interactions among the variables. Data were collected from 97 project leaders who have worked on government projects within the Cape Coast Metropolis. The study found that project policies and transformational leadership styles enhance project outcomes. On the contrary transactional leadership style was found to adversely impact project outcomes. The study further concludes that leadership style does not moderate the relationship between project policies and project outcomes within the CCMA. It is, therefore, recommended that clear policies be formulated to guide the execution of government projects while project leaders are trained and admonished to employ the transformational leadership style to increase the success rate of government projects.


PUBLIC INTEREST STATEMENT
This study was premised on the ground that a number of projects continue to achieve its outcome however, the same cannot be said about Ghana which continuously records a number of project failures.However, all these projects are expected to have a project policy which provides details on the rules, structures, protocols, relations and systems.Is it always the case that when projects have a project policy it leads to the achievement of the projects et targets?To explain this, the study indicated that not only will projects realize their intended targets but to achieve that project leaders must exhibit certain leadership styles.The transactional and transformational leadership styles were used to explain that the achievement of project outcomes when the project has developed project policy could be enhanced or impaired depending on the leadership styles that project leaders exhibit.

Introduction
Project outcomes have recently received tremendous attention in the project management literature.Contemporary projects are intensely planned, strongly implemented and thoroughly observed using modern management techniques and innovative technological tools; yet about 60% of these projects fail on meeting their expectations (Taherdoost & Keshavarzsaleh, 2016).According to the Project Management Institute (PMI) (2019), in 2018, only about 70% of the total projects were able to meet their original project goals, meanwhile only 60% of those projects were completed within the original budget.Iqbal et al. (2019) further surmised that project failures are reported around the world more than successful projects.
Particularly for Ghana, it is reported that the rate at which projects fail is high and alarming, with one out of every three infrastructural development projects failing or being challenged to achieve either of the three constraints of cost, time and scope (Amponsah, 2013;Damoah, 2015).Ofori-Kuragu et al. (2016) also suggested that most projects in Ghana do not meet their expected targets of time, cost and quality set by the project stakeholders and the outcome of these projects are also sometimes unsatisfactory (El Emam & Koru, 2008;Kappelman et al., 2006).According to Amoatey et al. (2015), the contribution of government construction projects cannot be underestimated as it serves as the basis for the rest of the economy to flourish.When government projects fail, it stalls the development agenda of the country.
Practitioners and scholars alike continue to investigate the factors that impact project outcomes.Several factors have been revealed to impact projects such as goal clarity, user engagement, management support, realistic expectations, time management, authority delegation, effective leadership, change management, project risk and quality, proper planning and project complexity (Durmic, 2020;Gemino et al., 2020;Handzic, 2017;Luo et al., 2017;Moohebat et al., 2010;Nasir & Sahibuddin, 2011).Obviously, most projects fail to succeed due to a lack of organizational culture and leadership instead of poor management and planning (Al-Qubaisi, 2015).More so, extant literature in the last four decades has highlighted a variety of project success factors even though the success criteria also keep increasing (Durmic, 2020;Josline & Müller, 2016).
Corollary to this, scholars have begun to broaden the scope of possible project success factors and focus more on the structural characteristics of the project context and its impact on success.One factor that is evolving in the literature is the project policy's impact on project outcomes (Biesenthal & Wilden, 2014).Müller and Lecoeuvre (2014) identified the structural characteristics needed for successful project execution, highlighting project policy as one of them.The nascent literature has also highlighted the influence of project policies and governance on project outcomes (Bekker & Steyn, 2008;Hirschey et al., 2009).
In this context, project policy is a policy guideline or document that provides the framework, rules, structures, protocols, relations and systems within which decisions are made throughout the lifespan of the project to accomplish the objectives of that project (Delhi & Mahalingam, 2020).It is "the use of systems, structures of authority, and processes to allocate resources and coordinate or control activity in a project" (Pinto, 2014, p. 14).Project policies also delineate how uncertainties arising from the execution of the project would be dealt with.Furthermore, they also mediate and provide mechanisms on how divergent views and issues on a project must be addressed (Li et al., 2019).
Nonetheless, these policies may also put some constraints in the way of the project manager/ leader during the project implementation and execution phases.Most of these policies are developed before the project leader is selected and little or no input is taken from him/her, which may further hinder his/her performance (Josline & Müller, 2016).In as much as the project policy can constrain a project manager or team, an effective leader can manoeuvre his or her way around the project policy to bring the best out of the project.The argument here is that the leadership style of the project manager could be a plausible difference between how the project policy affects the outcome of the project.For instance, the leader could deploy his or her leadership style to lure the followers to give off their best, even under the dire constraints of the project policy.
Extant literature suggests that effective leadership is a key contributor to successful project delivery and can therefore not be ignored in everyday life as well as in projects (Banihashemi et al., 2017).Leadership on a project level is more complicated than at an organizational level and therefore the leader's style is germane to the outcome of the project.Project employees are also bonded for a temporary time in a limited scope.Therefore, the least motivated followers can lose focus within that period (Rehman et al., 2020).Many authors have realized the need for further research on project leadership for both theoretical and empirical reasons.They emphasized that leadership style and personality traits are salient to look at in determining the outcome of a project (Anantatmula, 2010;Geoghegan & Dulewicz, 2008).Damoah and Kumi (2018) provide evidence by reporting that project leadership is one of the top ten determinants of project outcomes in Ghana.The leadership literature in contemporary project settings suggests that the most prominent styles of leadership are the transformational and transactional leadership styles (Bass, 1985;Northouse, 2014).These leadership styles come with their own merits and demerits and suit different situations and contexts to influence the outcomes of projects.Consequent to this, we contend that, apart from the project policy's impact on a project, the style of the leader is also paramount for the success of the project.
Generally, a few studies have investigated the impacts of project policies (governance) on project outcomes, albeit in foreign jurisdictions (Bekker & Steyn, 2008;Durmic, 2020;Hirschey et al., 2009;Josline & Müller, 2016).However, the literature on how project policies impact the outcome of projects, especially in Ghana, is very scanty and limited.Furthermore, the impact of the project leader's style on government project outcomes has also not received much attention in the Ghanaian context.As a novelty, this study employs two contemporary leadership styles (transformational and transactional leadership styles) to examine the moderating role of project leadership styles in the relationship between project policy and project outcomes in Ghana using government projects.

Theoretical framework
This study was grounded on the theoretical propositions from the Resource-Based View (RBV) and the Situational Leadership theories.The RBV theory contends that resources are scarce; nevertheless, they are the key drivers of competitive advantage, specifically in the project management context (Almarri & Gardiner, 2014;Wernerfelt, 1984).This theory is deployed in this research since it concentrates on resources (which involve leadership capabilities in projects and project policies as well) and how it influences a firm's performance (project outcome).The study focuses on how strategic resources such as project policies and leadership styles influence the outcome of government projects.Similarly, situational leadership theory posits that a leader exercises a particular leadership style based on the current circumstance (Hersey et al., 1979).Linking the theory to this study, it is realized that the project manager observes the different forms of construction projects and their policies which serves as a guiding principle before adopting a particular leadership style.The nature of the policies within a project organization may and can determine the type of style the project leader adopts.A clearly stated project policy calls for minimum supervision by a project leader whilst a vague stated policy would call for close supervision by a project leader (Delhi & Mahalingam, 2020).It is contended that the causes of project success or failure and its subsequent effects partly rely on the project policies and leadership style adopted in managing the project.Figure 1 presents the conceptual model as informed by the literature review.Based on the literature reviewed, the study formulated the model to guide the study in addressing the study's objectives and hypotheses.The arrows in the conceptual model explain the interrelationships proposed among the variables.The model indicates that having a project policy contributes to project outcome, which is defined by the direction of the arrow from project policy to project outcome.the study further proposed that both transformational and transactional leadership moderate the relationship between project policy and project outcome.The direction of the arrow from both transactional and transformational leadership to the main arrow moving from project policy to project outcome indicates that both transactional and transformational are moderating variables.

Project policy and project outcome
According to Damoah (2015), the outcome of a project is typically regarded in terms of whether the project has been successful or failed.Detecting what constitutes "success" or "failure", nevertheless, can be a challenging one.Project outcome can be defined within the remit of what establishes the project outcome (Agarwal & Rathod, 2006), who evaluates the project (Berssaneti & Carvalho, 2015), the criteria used in evaluating the performance of the project (Mir & Pinnington, 2014) and the timing of the evaluation (Heeks, 2010).In managing projects, there is a need for a clear demarcation of guidelines within which the objectives of the project can be evaluated (Müller, 2017).These guidelines are contained in a project policy document.The policy provides the framework for stakeholder management and decision-making during the lifecycle of the project (Josline & Müller, 2016).The policies serve as guidelines in the implementation of projects and can contribute greatly towards the successful outcome of a project.Kamau et al. (2020) also surmised that when there is a clearer interpretation of project policies in the construction industry, it enhances project performance.Consequent to this, we hypothesize that: H1: Clear project policies positively enhances the outcome of a project.

Transformational leadership style and project outcome
The project manager's transformational leadership style is paramount to the outcome of the project.Transformational leadership is about igniting the fires of people's inspiration and imagination (Oh et al., 2019).Zaman et al. (2019) explained transformational leadership as one that enunciates a shared vision, enhances followers intellectually and identifies the differences between employees.This type of leadership looks at enhancing employee motivation and engagement and attempts to connect their sense of self with organizational values (Fitzgerald & Schutte, 2010;Rao, 2014).This leadership style facilitates team building among project team members which in turn leads to project success (Aga et al., 2016).Prabhakar (2006) found the transformational leadership style to positively influence the outcome of 153 projects across 28 countries.Miyamoto (2015) also reported a direct relationship between transformational leadership and ICT Projects in Japan.Maqbool et al. (2017) further surmised that transformational leaders are more effective on projects than transactional leaders.Similar results were also reported in other studies such as Raziq et al. (2018), Zaman et al. (2019) and Oh et al. (2019).Based on the foregoing, we hypothesize that; H2: Transformational leadership has a significant positive effect on project outcomes

Transactional leadership style and project outcome
Transactional leadership involves an exchange procedure which is based on the accomplishment of contractual obligations, typically signified by setting objectives and monitoring and controlling the outcome (Aga, 2016).This approach to leadership enhances compliance with the already existing performance expectations and organizational goals through the use of reward and punishment and direct supervision (Northouse, 2014).Such leaders seek to stimulate their followers by alluding to their self-interests.These types of leaders encourage followers to understand the required performance levels by helping the followers to recognize their tasks and responsibilities, develop the confidence to perform up to standard and help followers identify goals (Aga, 2016;Lo et al., 2010).Regarding project organizations, the project leader first clarifies the project to the followers, informs them of their roles and the expected outcome and closely monitors the activities of the followers.Existing studies have documented a direct impact of transactional leadership on project success.For instance, Raziq et al. (2018) opined that the transactional style of leadership is positively related to project outcomes.The findings of Oh et al. (2019) also depicted that the transactional leadership style is more efficient in delivering favourable project outcomes in the public sector.corollary to these, we contend that; H3: Transactional leadership has a significant positive effect on project outcomes

Project policy, project leadership and project outcome
In a typical project organization, the project policies would greatly dictate how the project is to be implemented.However, the leadership style adopted by the project leader can influence how the policies affect the project.The leader could deploy his or her leadership style to lure the followers to give off their best, even under the dire constraints of the project policy.We argue that leadership style can be a catalyst in the nexus between the project policy and the outcome of the project.The literature is sparse about the association among these variables.However, few studies have partially examined the direct relationships among project policies, leadership and the outcome of projects.For instance, Aga et al. (2016) tested the moderating role of one dimension of project policy (goal clarity) in the relationship between project leadership and outcomes and documented that transactional leadership has more pronounced effects on project success when there is goal clarity.Similar evidence was found by Raziq et al. (2018), who concluded that goal clarity partially moderates the relationship between project leadership and project outcomes.In line with these arguments, we hypothesize that: H4: Transformational leadership moderates the relationship between project policies and project outcomes H5: Transactional leadership moderates the relationship between project policies and project outcomes

Research methods
The survey data was used for the analysis and as was indicated in the study of Meier and O'Toole (2013), surveys are best when researchers are interested in how managers operate.Since the current study sought to address similar issues, the survey was deemed appropriate.The survey data were collected from project leaders who had been engaged by the government to undertake projects within the Cape Coast Metropolitan Assembly (CCMA).The study used a time frame of ten years between 2010 and 2020.This period was deemed long enough for projects that were initiated within that period to have successfully gone through the complete life cycle.Further, several government projects had been undertaken during the period within the study area.The CCMA revealed that within this period, 137 project managers had been engaged in government projects for which their complete contact data were available at the assembly.The 137 project managers were, therefore, employed as the population for this study.However, only 110 of these personnel consented to participate in the study.Considering the number of project managers that consented to participate in the study, the census approach was used.
To adhere to COVID-19 safety protocols and guarantee a higher response rate, those who agreed to participate in the study were emailed a structured questionnaire in a Google form document format.Measures of project policy were adopted from Josline and Müller (2016) and Ul Musawir et al. ( 2017) while the measures for transformational and transactional leadership style were made up of 10 items each adapted from Northouse (2014), Dartey-Baah and Ampofo (2015) and Yukl and Mahsud (2010).According to Avolio (1999); Burns (1978), the more effective and most effective leadership styles are transactional and transformational leadership respectively, hence the focus of the study in limiting the leadership styles to these two.Finally, 13 items were adopted from Al-Qubaisi (2015) and Damoah and Kumi (2018) to measure project outcomes.A total of 97 participants responded to the questionnaire, giving us a response rate of 88.18%.To address the self-report bias, which is the tendency of the respondents to be biased or subjective in their responses, the study was guided by Meier and O'Toole (2013).This ensured that the selfreport bias was reduced to ensure the accuracy of the data.

Data processing and analysis
The Partial Least Squares-Structural Equations Modelling (PLS-SEM) approach examines the relationships between the variables.The Smart PLS version 3.3.3software was employed in the analysis to obtain the path coefficients, average variance extracted (AVE), composite reliability (CR), bootstrap results and all other results presented in the next section.According to Adam et al. (2017), the PLS approach uses an iterative process of principal component analysis (CPA) and regression analysis.The benefit of this approach is that the results are estimated simultaneously to avoid issues of parameter inconsistencies and biased estimates.The PLS approach also places less emphasis on sample size, and measurement scale and allows for several regressors (Benitez et al., 2020;Hair et al., 2019;Ringle et al., 2020).

Results and discussions
We began by evaluating the measurement and structural models of the PLS-SEM results to verify the reliability and validity of the constructs and also assess the relationship between the constructs.

Measurement model
In assessing the measurement model for a reflective scale, Hair et al. (2019)  loadings range from 0.709 to 0.862, implying that indicator reliability has been met.Next, we check the model for internal reliability using Cronbach alpha, Rho A and composite reliability.Reliability values above 0.7 are considered ideal.Again, the results presented in Table 1 show that all three measures of reliability meet the acceptable criteria of 0.7 for all the indicators.Internal reliability is therefore confirmed for all the variables in the study.To examine the model for convergent validity, the Average Variance Extracted (AVE) values are used.Ringle et al. (2020) propose that this measure should exceed 0.5 to denote the presence of convergent validity.Table 1 confirms that all the constructs have AVE values ranging from 0.584 to 0.684.this suggests that all the construct explains more than 50% of the indicator's variance.Finally, the study assessed the model for discriminant validity using the HTMT ratio.Hair et al. (2019) suggest that the HTMT ratio is preferable to the Fornell-Larcker criterion because the Fornell-Larcker performs badly when the item loadings on each construct only differ slightly.Ringle et al. (2020) recommend a threshold of 0.85.Discriminant validity is therefore confirmed as all the HTMT values reported in Table 2 are below the 0.85 threshold.The Variance Inflation Factors (VIF) were also below the recommended threshold of 3.3 to 5, as suggested by Hair et al. (2019).

Assessment of structural model
After confirming the measurement model, we further assessed the structural model for the statistical significance of the constructs and the predictive power of the model.The results, as captured in Figure 2 and Table 3, depict that all the beta coefficients in the model without the moderating variables were statistically significant at 1%.However, the model with the moderating variables depicts that the path betas for PP*TF>PO and PP*TS>PO were not statistically significant.
More so, the model depicts a strong predictive power.85% of the variance in project outcomes is explained by the three latent variables, i.e. project policies, transformational leadership and transactional leadership.The study's predictive power was also achieved by the Q square of 0.429, which suggests that the model has a medium predictive relevance (Hair et al., 2019).Finally, the effect size (f 2 ) results suggest that project policy is the highest predictor of project outcomes among the exogenous variables.

Discussions
This study assessed the impact of project policies and project leadership on government project outcomes in Ghana.Five hypotheses were formulated to examine the impact of project policies (H1), Transformational leadership (H2), Transactional Leadership (H3) and the moderating role of the two leadership styles on the project policy project outcome nexus (H4 and H5).
The findings from the PLS-SEM analysis reveal that project policies significantly influence the outcome of government projects.With a positive coefficient of 1.661 (t-stat = 5.079; p-value <0.05), the results revealed that project policies have a statistically significant impact on the outcomes of projects within the CCMA.The finding therefore supports hypothesis H1 which states that clear project policies positively enhance the outcome of a project.This is practically so because, when the policies about the scope, objectives and goals of the project formulated by the project organization are clear and concise, all stakeholders will be well informed about the direction and purpose of the project (Raziq et al., 2018), which is likely to improve the outcome of the project (Aga, 2016;Anderson & Stritch, 2016).The results from the findings are in conformity with that of Aga ( 2016) who concluded that when there is goal clarity from project policies, it enhances the success rate of the project.Similar findings were also reported by Ul Musawir et al. ( 2017) who found effective project governance to positively influence all dimensions of project success.
The results further suggest that transformational leadership has a coefficient of 0.398 (t-stat = 3.650; p-value <0.05), signifying a significant positive effect on project outcomes.This means that ceteris paribus when the project leader's style is more transformational, the project's outcome is more likely to be positive (Zaman et al., 2019).The findings of the study imply that, in projects where the leader's style is more inspirational and transformative, team members are induced to produce a desirable and positive outcome on the project, which renders it into success (Aga et al., 2016).Consequent to this, hypothesis H2 is also supported.This is because, transformational leaders motivate and inspire team members to give out their best, while the leader takes care of the team members' needs and wants (Raziq et al., 2018).The leader promotes the interests of the team members along with the projects' objectives.This not only remove the hurdles a team member faces but also builds an environment of satisfaction, creativity, trust and accomplishment within the project team (Maqbool et al., 2017).This is likely to translate into the success of the project.This finding is overwhelmingly supported by the existing literature.For instance, Maqbool et al. (2017), Zaman et al. (2019) and Oh et al. (2019) have all found transformational leadership to have a more positive influence on project success.
From Figure 3 and Table 3, the results further revealed that transactional leadership style has an adverse influence on project outcomes, showing a negative coefficient of 1.128.The effect is statistically significant at 5% (t-stat = 3.109; p-value <0.05).The results imply that the transactional leader's impact on the outcome of a project is negative and therefore hampers the success of the project.
Even though this finding is contrary to the expectation of the researcher, it supports hypothesis H3.Transactional leadership is generally said to be an exchange process between the leader and team members, which is liable to a reward and punishment relationship (Oh et al., 2019).This leadership style focuses more on routine processes, compliance with organizational goals and performance expectations within the status quo (Caillier & Sa, 2017).Such leaders are more likely  "to think inside the box when solving problems and maintain the status quo" (Odumeru & Ifeanyi, 2013).This can be demotivating to project team members as individual creativity and innovation are stifled, leading to an adverse impact on project outcomes (Caillier & Sa, 2017;Mirza & Javed, 2013).Maqbool et al. (2017) therefore concluded that team members rate transformational leaders to be more effective than transactional leaders.

Regression with moderation
The study further tested the moderating role of each of the leadership styles in the relationship between project policy and project outcomes by including an interacting variable of project policy and transformational leadership (PP*TF) and transactional leadership (PP*TS) in the model.At the introduction of the moderating variables into the model, the results changed slightly.It is observed that project policies consistently had a positive impact on project outcomes even though the coefficient had slightly improved (β = 1.717, t-stat = 4.337; p-value <0.05) in line with H1.
Transformational leadership was also positive and significant (β = 0.360; t-stat = 3.338; p-value <0.05), as was reported earlier in support of h2.It is also observed that the effect of transactional leadership has slightly declined as compared to the one without the moderator (β= −1.168; t-stat = 2.741; p-value <0.05).
Contrary to the argument of the researchers that project leadership can condition the effect of project policies on project outcomes, the results proved otherwise.The model shows that PP*TF and PP*TS have a positive and negative coefficient of 0.059 and −0.099 respectively.However, both variables' contribution to project outcomes is statistically insignificant.We therefore reject H4 and H5 respectively.The results suggest that either of the leadership styles do not moderate the relationship between project policies and project outcomes within the CCMA Metropolis.However, the positive and negative coefficients of transformational and transactional leadership styles are in conformity with Bass's and Bass (2009) conclusion that transformational leaders are more effective than transactional leaders in project execution.

Conclusions and recommendations
This study sought to investigate the relationships among project policies, leadership and project outcomes in Ghana, with an emphasis on government projects.Corollary to the findings from this study, a number of conclusions were arrived at.First and foremost, the study concludes that project policies that are clear to all team members enhance the execution of the project and hence translate into the overall success of the project.In line with the RBV theory, the study concludes that project policies are strategic resources an organization can use to enhance the success of its projects.More so, the study's findings support extant literature on the motivational impacts of a transformational leader on project outcomes and therefore conclude when managers of a inspire their team members, it contributes towards the attainment of project outcomes.On the contrary, when project leaders adopt a transactional style, it limits their engagement with project team members thereby affecting the attainment of project outcomes.Based on the situational leadership theory, we conclude that a transformational leader is more effective in project settings than a transactional leader.The evidence further suggests that leadership style does not moderate the project policy project outcome nexus within CCMA projects, implying that irrespective of the style of the project leader, he or she is not able to change the impact of project policies on the projects' outcome and for that matter, We, therefore, recommend that project policy formulation be handled strategically by the formulating authority.This must be done based on broader consultations to factor in the experiences and opinions of all stakeholders to enhance the effects of the policies on projects.In instances where certain aspects of a policy might have outlived their relevance, they should be amended or changed entirely to harness the full benefits of these policies on the projects executed within the metropolis.
The study further recommends that irrespective of the leadership style of the leader, he or she must place greater emphasis on clarifying and explaining the project policy requirements, goals and other expected standards about the project, since if that is not done, the outcome of the project is likely not to be favourable.The study also recommends that, in the best interest of the project's outcome, a dynamic approach to leadership be adopted by the project leader, with greater emphasis on a transformational style.Project organisations can also admonish their project leaders to employ the transformational style as they execute various projects.

Limitations of the study
The study employed 97 project leaders engaged in government projects within CCMA.This can affect the generalizability of the study findings to include all project leaders in the region.The study concentrated on the Cape Coast Metropolis and the findings are only applicable to this context.Again, factors such as budget, project complexity and other confounding variables were not included in the analysis.Finally, the project leaders evaluated their own style of leadership but the views of the team members were not considered to provide a 360-degree view.It is therefore likely that social desirability bias would be introduced by the respondents.Based on these, further studies can expand the study to include other confounding variables such as budget, political or economic factors, and stakeholder engagement to determine how these variables influence the relationship between project policy, project outcome and leadership styles.Further, since the study was a cross-sectional study, providing a holistic view of the leadership style used by project managers over time requires a longitudinal study.Future studies can therefore consider undertaking a longitudinal study to address this challenge.

Table 1 . Assessment of the measurement models Construct Items Outer loadings Cronbach alpha Rho A Composite Reliability AVE VIF
recommend the evaluation of the outer loadings, internal consistency reliability, convergent reliability and discriminant validity.As captured in Table 1, all indicators had factor loadings greater than the 0.709 that was recommended by Ringle et al. (2020).It is observed from Table 1 that the indicator