The Double-Edged Sword of Integrity: How Ethical Compliance Attenuates the Capability–Performance Link in Ghana’s Local Government Projects

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for taking this issue up for research, it is interesting to read the relationship between capability-ethical consideration-performance. Paper has a lot of potential but it requires through revision. Below are my suggestions:

1. Introduction section is well structured however hypotheses must not be presented in the this section.
2. Discuss the potential so what part in the introduction section itself. 
3. Literature review needs a theory section that would discuss the relevant assumptions of theory and how this research is based in them. 
4. Also add more to discussion section that how this is adding to extant knowledge base. 

Author Response

See attached file

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript investigates how project team capability and ethical compliance influence project performance in Ghana’s public sector, with a focus on the moderating role of compliance in the capability–performance link. The topic fits well within the scope of Administrative Sciences, since it addresses public management, governance, and organizational performance in administrative settings.

 

Comment 1: The section of introduction effectively situates public project performance within governance and ethical compliance; however, its structure makes it difficult to immediately identify the research gap, the central research question, and the manuscript’s scholarly contribution.

C.1.1 Lines 112–124 state that “it is not sufficiently clear how ethical compliance interacts with project team capability,” yet the claim remains broad and does not specify which aspect of this interaction is underexplored in prior research.

C.1.2 Lines 125–128 indicate that capability and compliance have largely been studied “as separate,” but the manuscript does not clearly establish why examining them together constitutes a substantive theoretical gap rather than an incremental extension.

C.1.3 Lines 145–152 present a sequence of research objectives without articulating a single, clearly identifiable research question, which gives the introduction a descriptive rather than analytically driven orientation.

C.1.4 Hypotheses H1–H4 are introduced in lines 160–186 before the theoretical development is completed, disrupting the expected logical progression from theory and literature toward hypothesis formulation.

C.1.5 Lines 57–96 contain an extended discussion of ethical compliance mechanisms and governance dynamics that reads primarily as literature background, thereby shifting the introduction away from clearly positioning the manuscript and its contribution.

Comment 2: This section (Theory and Hypotheses Development) demonstrates solid theoretical grounding by integrating RBV and Ethical Leadership Theory and linking them to project capability, compliance, and performance. However, the narrative often becomes descriptive, and the analytical bridge between theory, constructs, and hypotheses is not always sufficiently tight, which slightly weakens the theoretical positioning of the manuscript.

C.2.1 In lines 274–286, RBV is introduced mainly at a conceptual level, but the manuscript does not clearly specify which attributes of RBV (e.g., VRIN characteristics) directly justify project team capability as a strategic resource within this specific model.

C.2.2 Lines 296–306 describe Ethical Leadership Theory broadly, yet the mechanism through which ELT explains ethical compliance as a structural governance control remains implicit rather than theoretically articulated.

C.2.3 In lines 328–341, mixed empirical findings are acknowledged regarding the capability–compliance relationship, but the manuscript moves directly to hypothesis formulation without clearly resolving this inconsistency at the theoretical level.

C.2.4 The moderation logic (lines 372–388) introduces ethical compliance as a “contextual force,” yet the theoretical boundary conditions under which compliance constrains capability are not clearly specified, leaving the moderation rationale somewhat underdeveloped.

C.2.5 The hypotheses across sections 2.5.1–2.5.4 largely restate expected relationships from prior literature, but the manuscript does not consistently distinguish whether these hypotheses test replication, extension, or theory-building, which makes the theoretical contribution less explicit.

Comment 3: The methodology is generally well structured, clearly presenting the research design, sampling approach, measurement scales, and PLS-SEM procedure. However, several elements related to data origin, sampling logic, and analytical justification require greater precision to ensure methodological transparency.

C.3.1 Lines 421–423 state that the dataset is part of the corresponding author’s PhD research and has been used in other papers with alternative models, yet the manuscript does not clarify whether the present study relies on the same respondents or measurements, which raises potential concerns regarding data independence.

C.3.2 The sampling strategy is described as purposeful (lines 418–420), but the criteria used to identify and select eligible senior officers are not specified, making it difficult to assess the rigor and potential selection bias of the sample.

C.3.3 PLS-SEM is presented as suitable for “theoretically driven models with multiple latent constructs and smaller to medium-sized sample sizes” (lines 441–444), yet the manuscript does not explicitly justify its preference over covariance-based SEM given the relatively robust sample of 320 responses.

Comment 4: The results section is logically organized and reports measurement validity, reliability, and structural model outcomes in a clear sequence. However, several reporting inconsistencies and interpretive elements reduce analytical precision and blur the boundary between results and discussion.

C.4.1 In Table 3 (lines 500–503), Cronbach’s alpha for Team Capability is first reported as 0.847 and later described as .92, creating an internal inconsistency in reliability reporting.

C.4.2 Lines 535–538 interpret an R² of 0.314 as “moderate explanatory power” but describe 0.018 as “extremely low,” without referencing accepted thresholds, which makes the evaluation appear subjective rather than statistically grounded.

C.4.3 Table numbering is duplicated (“Table 5” appears twice around lines 533–544), which may create confusion when cross-referencing results within the manuscript.

C.4.4 Lines 490–496 characterize construct validity as “strong” and “robust,” introducing evaluative language within the results section instead of limiting the text to statistical reporting.

C.4.5 The moderation finding is described as both having a “small effect size” (lines 547–549) and representing a “moderate degree of moderation,” producing ambiguity regarding the practical significance of the interaction effect.

Comment 5: The discussion engages extensively with prior literature and connects the findings to RBV and ELT; however, parts of the section lean toward literature reiteration rather than critical interpretation of the study’s results. Key elements expected in a discussion—clear hypothesis confirmation, sharper theoretical contribution, and structured implications—are present but not always distinctly articulated.

C.5.1 Lines 589–600 restate prior empirical studies supporting the capability–performance link, but the discussion does not clearly differentiate what the present findings add beyond confirming established RBV arguments.

C.5.2 Although all hypotheses are statistically supported earlier, the discussion does not explicitly synthesize their confirmation into an integrated interpretation of the model, leaving the overall theoretical message somewhat fragmented.

C.5.3 Lines 635–646 list multiple theoretical contributions; however, several points reiterate RBV assumptions rather than specifying how the study advances, refines, or challenges existing theory.

C.5.4 The practical implications (lines 680–705) are predominantly normative (e.g., emphasizing training and capability development) and are not consistently anchored to the moderation finding, which appears to be the manuscript’s most distinctive result.

C.5.5 Despite referencing governance trade-offs, the discussion does not critically address the relatively modest explanatory power reported earlier (e.g., R² = 0.314), nor does it reflect on how this affects the strength or generalizability of the conclusions.

Comment 6: The conclusion summarizes the interplay between capability, ethical compliance, and performance in a coherent manner; however, parts of the section read as a thematic recap rather than a sharply delimited closure that consolidates the manuscript’s scholarly contribution.

C.6.1 Lines 708–717 restate that performance emerges from the alignment of team skills, ethical climates, and governance arrangements, yet the conclusion does not explicitly articulate what this alignment demonstrates at the knowledge level (e.g., whether it reframes capability theory in public project settings or clarifies the role of compliance as a structural condition).

C.6.2 The limitation paragraph (lines 721–730) identifies constraints related to structured questionnaires, single respondents, and quantitative design, but it does not specify how these limitations may have influenced the observed relationships, leaving their methodological implications insufficiently defined. Pls move limitations to discussion section.

Author Response

See attached file

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors,

Your manuscript tackles a genuinely interesting and underexplored tension—the possibility that well-intentioned ethical compliance regimes may inadvertently blunt the performance edge of capable project teams. However, the current execution falls significantly short of the rigour required for publication. The core problem is not your research question, but the chasm between what you claim and what you actually deliver. You claim a national study of Ghana’s public sector, yet your data is drawn from a single region. You claim strong psychometric properties, yet your ethical compliance scale has a Cronbach’s alpha of 0.653 and you retained low-loading items based on post-hoc rationalisation. You claim a meaningful moderating effect, yet the effect size (F² = 0.020) is vanishingly small. You claim to have addressed common method bias, yet you used a single-source, cross-sectional survey with no procedural remedies and relied on a post-hoc statistical test as a fig leaf. The duplication of Table 5 is not a typo; it is evidence of a manuscript prepared in haste. This paper has a viable core, but it is buried under overclaiming, conceptual blurring, and methodological corner-cutting. Please see my comments below and make the necessary corrections:

Title

The title is functional but lacks scholarly gravitas and precision. It announces a "moderating effect" but buries the core theoretical contribution—the "control-flexibility trade-off"—which is not signposted. This is a missed opportunity to frame the paper’s counterintuitive contribution. Furthermore, the geographic qualifier "Evidence from Ghana's public sector" is relegated to a subtitle, yet the study is actually confined to the Volta Region’s MMDAs. This is not merely a semantic quibble; it is a fundamental misrepresentation of scope. The title promises a national study but delivers a regional convenience sample. This overclaiming must be rectified. Either the title should specify "Volta Region" or the authors must concede that generalising to "Ghana's public sector" is unwarranted. I recommend: "The Double-Edged Sword of Integrity: How Ethical Compliance Attenuates the Capability-Performance Link in Ghana’s Local Government Projects." This reframes the negative moderation as the central plot, not a secondary finding.

 

Abstract

The abstract is verbose and structurally incoherent. It begins with a normative statement about what public-sector projects must do, which is editorialising, not social science. The phrase "it remains unclear whether ethical compliance simply acts as a direct performance enhancer, or whether it also shapes how effectively project teams convert their skills into outcomes" is a false dichotomy. The study tests both, and finds both. The abstract fails to clearly prioritise which finding is the novel contribution. The negative moderating effect is mentioned, but its magnitude (β = -0.107) and practical significance are omitted. The reader is told there is a trade-off but not how severe it is. Moreover, the abstract claims the study “conceptualizes ethical compliance as an integrity control architecture,” yet this conceptualisation is not operationalised or measured as such. It is a post-hoc labelling of a latent variable measured by three items with borderline reliability (α = 0.653). The abstract oversells the theoretical innovation. Finally, the policy prescription (“proportionate, risk-based compliance regimes”) is unsupported by the data, which contains no measures of regulatory design or proportionality. This is a normative leap, not an empirical implication.

 

1. Introduction

The Introduction suffers from academic throat-clearing and a failure to establish a clear, narrow problematisation until very late (Page 4, lines 145-160). The opening paragraph reads like a generic textbook on public sector reform, citing eight references in the first two pages without advancing a specific argument. The authors invoke “governance reforms,” “corruption,” “stakeholder trust,” and “public sector scrutiny” as a sprawling laundry list of concerns. This is not a precise research problem; it is a literature review in miniature.

Crucially, the Introduction misrepresents the state of the literature. It claims that “interactive effects are still insufficiently evidenced in developing-country public sector delivery settings” (lines 142-144). This may be true, but the authors do not cite a single systematic review or meta-analysis to substantiate this claim. It is a convenient gap statement, not a demonstrated lacuna. Furthermore, the authors do not engage with the rich literature on regulatory burden, administrative burdens, or red tape in public administration (e.g., work by Bozeman, Pandey, Moynihan). By framing the negative moderation entirely within project management and construction journals, they miss the opportunity to connect their finding to a well-established public administration paradigm. This is a serious theoretical omission.

The research objectives (i-iv) are not presented as formal research questions but as a narrative list. This lack of formalisation weakens the scientific rigour of the exposition. The authors also state that “public organisations have no choice but to make ethical compliance central” (lines 76-77). This deterministic, agent-less language is inappropriate. Organisations make strategic choices; they are not passive victims of institutional scripts.

 

2. Theory and Hypotheses Development

This section is the weakest pillar of the manuscript. It purports to integrate Resource-Based View (RBV) and Ethical Leadership Theory (ELT), yet the integration is superficial—a patchwork, not a synthesis.

2.1 Conceptualising Public Project Performance: This subsection is tautological. It defines performance as “the balance between delivery targets and ethical implementation” (lines 216-219), yet this is precisely the phenomenon the study seeks to explain. The authors have built their dependent variable to contain the very tension they are testing. This is a classic case of conceptual stretching and risks tautology in hypothesis testing. If performance already includes ethical implementation, then regressing ethical compliance on performance is circular. This must be clarified: is ethical performance an input, a process, or an outcome? The measurement model must be purged of this conflation.

2.2 Project Team Capability: The definition is adequate but relies heavily on textbook sources (Sacks et al., 2018; Fewings & Henjewele, 2019). There is no critical engagement with competing frameworks of team capability (e.g., dynamic capabilities, absorptive capacity). The construct is treated as monolithic, yet the five items used to measure it (PTC1-5) tap into disparate dimensions: communication, technical skill, resource management, adaptability. The authors do not justify why these are formative or reflective, nor do they test for multidimensionality. This is a latent variable modelled as unidimensional, but the conceptualisation suggests otherwise.

2.3 Ethical Compliance as Integrity Control Architecture: This section is riven with definitional ambiguity. The authors conflate ethical compliance (adherence to rules) with ethical culture (shared values) and integrity systems (structural controls). These are distinct constructs with different nomological networks. The measurement items (ETH1, ETH4, ETH5) are not reproduced in the manuscript, but the references to Adeniyi et al. (2020) and Sarawa & Masud (2020) suggest a focus on procurement compliance and anti-corruption measures. This is a narrow operationalisation of “ethical compliance” that does not match the expansive theoretical definition provided (which includes “fairness, transparency, accountability, and respect”). The gap between conceptual definition and operationalisation is glaring and unacknowledged.

2.4 Theoretical Lens: The RBV-ELT pairing is forced. RBV explains why capability drives performance (VRIN resources). ELT explains why ethical leadership influences follower behaviour. But neither theory directly explains why compliance would weaken the capability-performance link. The authors resort to an ad hoc explanation: “compliance inhibits discretion” (lines 182-184). This is a plausible behavioural mechanism, but it is not derived from either RBV or ELT. It is borrowed implicitly from organisational control theory or street-level bureaucracy. The authors should either expand their theoretical repertoire or explicitly state that they are using RBV and ELT only for the direct effects, and a separate logic for the moderation. The current presentation is intellectually dishonest.

2.5.4 The Moderating Role: The hypothesised negative moderation is framed as a “central tension,” yet the language used to introduce it is tentative (“could potentially,” “may act as a boundary condition”). This indecisiveness is inappropriate for a hypothesis section. H4 is also incorrectly specified in the text. On page 8, lines 366-369, the authors write: “H4: ethical compliance moderates the relationship...” This is a non-directional hypothesis. However, in the initial presentation (page 4, lines 177-185), H4 is explicitly directional: “project ethical compliance has a negative moderating effect... becomes weaker.” This inconsistency suggests careless editing. A directional hypothesis must be stated as such from the outset.

 

3. Materials and Methods

This section reveals multiple fatal flaws that, in a truly rigorous review, would warrant outright rejection or mandatory major revision.

Sampling and Generalisability: The study is confined to 18 MMDAs in the Volta Region. This is a convenience sample, not a representative national sample. The authors claim this is “purposeful” sampling (line 416), but purposeful sampling requires strategic selection based on specific criteria (e.g., high-performing vs low-performing districts). No such stratification is described. This is a convenience sample of accessible units. The claim that this context is “suitable” is insufficient justification. The response rate (74.1%) is commendable, but this does not mitigate the lack of representativeness.

Common Method Bias (CMB): The authors collected all data from the same respondents using the same instrument at the same time. This is a cross-sectional, single-source design. The authors report that “all VIF values were less than 3.3” and conclude “no concerns with either of these issues” (lines 457-459). This is a fundamental misunderstanding of the Harman/VIF approach. The full collinearity test (Kock, 2015) is a post-hoc statistical diagnostic, not a license to ignore design flaws. The study lacks any procedural remedies for CMB (e.g., temporal separation, psychological separation, multi-source data). The conclusion that CMB is “not a concern” is scientifically indefensible.

Reliability: Table 3 reports Cronbach’s alpha for ethical compliance as 0.653. This is below the conventional 0.70 threshold. The authors acknowledge this but do not problematise it. They simply state “the results demonstrate strong psychometric properties” (line 516). This is an outright misrepresentation. An alpha of 0.653 indicates questionable reliability. The composite reliability (rho_c = 0.808) is marginally acceptable, but this does not excuse the low alpha. The authors must discuss the implications of measurement error on their estimates.

R-squared for H2: Table 5 shows that project team capability explains only 1.8% of the variance in ethical compliance (R² = 0.018). This is a vanishingly small effect. Yet the authors proceed to interpret the significant beta (0.134, p = 0.025) as substantively meaningful. With an R² this low, the statistical significance is driven almost entirely by sample size. The practical significance is negligible. This finding should be de-emphasised or re-framed as null. The authors’ discussion (page 16) inflates this trivial association.

Table 5 Duplication Error: The manuscript contains a critical typographical error. Table 5 is presented twice—once on page 13 (R-square) and again immediately below with the same title but now purporting to show F-square statistics (page 13, line 542). This is not a minor formatting issue. It indicates sloppy manuscript preparation and a failure to proofread. The editor’s confidence in the authors’ attention to detail will be justifiably shattered.

 

4. Results

4.1 Respondent Profile: The sample is 77.8% male and 53.4% postgraduate educated. This is a highly elite, male-dominated sample. The authors do not discuss the implications of this demographic skew for the generalisability of findings regarding “ethical compliance.” Do highly educated, male senior officers perceive compliance differently than junior, female, or less-educated staff? This is not explored. The homogeneity of the sample is a threat to external validity that is entirely ignored.

4.6 F-square (Effect size): The F² for the interaction term is 0.020. According to Cohen (1988), this is a very small effect size. The authors describe it as “moderate” (line 550). This is factually incorrect. An F² of 0.02 is small; 0.15 is medium. The authors are overstating their case. The negative moderation is statistically significant but practically trivial. The manuscript must be revised to reflect this accurately. The narrative currently implies the moderation is a major finding; the effect size suggests it is a marginal nuance.

Hypothesis Testing (Table 8): The path coefficient for H2 (TC -> EC) is 0.134, t = 2.238, p = 0.025. Given the low R² and the large sample size, this p-value is fragile. The authors should report confidence intervals and discuss the likelihood of replication failure. The coefficient for H4 (interaction) is -0.107. The authors do not provide a simple slopes analysis or a Johnson-Neyman plot. The reader cannot discern at what level of ethical compliance the relationship between capability and performance becomes non-significant or reversed. This is a major omission for a moderation study.

 

5. Discussion

The Discussion section reads as a victory lap, not a critical reflection. Each hypothesis is restated and “supported” by a string of citations that confirm the same pattern. There is no engagement with the null or contrary evidence within the study itself. The low R² for H2 is not mentioned. The marginal effect size for H4 is not contextualised. The lack of causality is acknowledged in one sentence (line 679) but then immediately ignored as the authors prescribe policy.

The theoretical contribution is overstated. The authors claim they “conceptualize ethical compliance as an integrity control architecture” (line 691), yet this is not a new conceptualisation; it is a relabelling of existing compliance literature. The claim of a “control-flexibility trade-off” is presented as a discovery, yet this trade-off has been extensively theorised in organisational sociology (e.g., Merton’s goal displacement, Selznick’s institutional defences). The authors do not cite or engage with this foundational work. Their contribution is incremental at best.

Practical Implications: The recommendation that “staff capability development programmes should bring together technical capability and ethics” (lines 702-704) is banal. It does not follow from the negative moderation finding. In fact, the negative moderation suggests that more compliance training for already capable teams may be counterproductive. The authors do not grapple with this uncomfortable implication. Instead, they retreat to safe, generic management consultancy advice.

Author Response

See attached file

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thanks for taking up this revision. The revised manuscript is much stronger overall, with a clearer focus. 

Author Response

No revieions required by this reviewer

Reviewer 2 Report

Comments and Suggestions for Authors

I would like to thank the authors for the careful revision of the manuscript. The manuscript has improved in clarity and overall coherence, and most of my previous comments have been adequately addressed. The study is now much clearer and better positioned.

Regarding Comment C.1.4.
In my previous review, I suggested reconsidering the placement of the hypotheses. In your response, you indicate that the hypotheses have been removed. However, in the current manuscript the hypotheses (H1–H4) still appear explicitly in the text. This creates a minor inconsistency between the response letter and the revised manuscript. Please clarify your intended approach and ensure full alignment.

Author Response

See attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

After reviewing the authors' revised version, I observed notable improvements and substantial changes. The authors have commendably addressed my previous ambiguities and feedback. The manuscript is now much improved, and I am pleased with these revisions.

Author Response

No revision required