Transformational leadership, sustainability governance, and organizational maturity in higher education institutions: a systematic review toward sustainable energy performance

Universities consume energy at an industrial scale, yet most struggle to translate sustainability commitments into operational change. Three constructs bear on this problem: transformational leadership (TL), sustainability governance, and organizational maturity. Each has its own research tradition, but whether and how they converge in the higher education literature remains unclear. This systematic review maps that convergence. Following PRISMAL 2020, we searched Scopus and Web of Science for work published between 2014 and 2026, identifying 230 records that were screened down to a final corpus of 34 studies for descriptive bibliometric and thematic analysis. Leadership was the most frequently central construct: 50.0% of studies treat it as a focus, followed by governance at 44.1%. Organizational maturity was central in only one study (2.9%). No study in the corpus treated all three constructs as central variables, and none paired leadership with maturity in the absence of governance; these patterns held under both strict (direct-only) and lax (direct-plus-partial) coding. The corpus was overwhelmingly unidimensional: 85.3% of studies addressed a single construct as a central variable. Asia and the Middle East contributed 55.9% of studies; Latin America only 5.9%, with a single study from Ecuador. Six thematic categories emerged, from green TL and PLS-SEM work to German governance frameworks and Latin American case studies. Within the reviewed corpus, two gaps are notable: no study examined how leadership builds organizational maturity for sustainability directly, and no study integrated all three constructs as central variables. Leadership maturity models exist in professional practice but were not identified in HEI sustainability research. Read as a research agenda rather than a settled conclusion, these patterns motivate five propositions oriented toward a Governance and Maturity Model (MGM) for scaling solar energy and efficiency projects on university campuses, with particular attention to Latin American contexts where structured maturity diagnostics appear most needed.

Universities run large physical campuses. They consume energy at a scale comparable to small towns, manage aging building stock, and face growing pressure to align operations with their own sustainability rhetoric (Leal Filho, 2019; Zen et al., 2016). The Sustainable Development Goals, particularly SDG 7 (affordable and clean energy) and SDG 13 (climate action), have sharpened that pressure (Bauer et al., 2021). The distance between what universities promise and what they deliver, however, remains wide (Rieg et al., 2021).

Three research streams address pieces of this problem. Transformational leadership (TL), built on Bass's (1985) four-factor model, has been linked to green behavior, staff commitment, and innovation in university settings (Kisahwan et al., 2025; Mukhtar et al., 2025; Farooq et al., 2024). Sustainability governance encompasses the formal and informal structures that universities use to coordinate sustainability across operations, teaching, research, and outreach, which have been mapped most thoroughly by German scholars (Bauer et al., 2018, 2020; Niedlich et al., 2020a) and studied in network form in Africa (Jimu and Rennkamp, 2024). Organizational maturity, rooted in capability maturity models (Humphrey, 1989), provides a diagnostic for assessing where an institution stands on a readiness continuum and what it should build next (Anthony and Antony, 2021; Niedlich et al., 2020b).

Each stream has produced useful findings on its own. The trouble is that they rarely talk to each other. Leadership without governance structures depends on individual champions and collapses with personnel turnover (Robinson and Laycock Pedersen, 2021). Governance without a maturity assessment generates policies that institutions cannot execute because no one has first diagnosed their capacity (Rieg et al., 2021). Yet, we are not aware of a review that has checked whether these three constructs converge in the higher education literature, particularly around campus energy performance.

Existing reviews have covered sustainability in higher education broadly (Leal Filho, 2019), organizational change for sustainability (Rieg et al., 2021), and sustainable leadership as a standalone topic (Liao, 2022). Within the literature we surveyed, we did not identify a study examining the three-way intersection of TL, governance, and maturity with energy performance as the outcome variable.

This literature gap is compounded by a theoretical one. Existing HEI sustainability reviews treat transformational leadership, governance, and maturity as adjacent literatures rather than as mechanisms that theoretically presuppose one another. Dynamic-capabilities theory frames organizational adaptation as a cycle of sensing opportunities and threats, seizing them through resource reconfiguration, and transforming the organization's asset base accordingly (Teece, 2007); read this way, leadership, governance, and maturity are not three parallel topics but three moments of a single capability-building process—a leader who senses an opportunity for change needs governance structures to seize it and a maturity trajectory to show that the change has been institutionalized. Absent a theoretical account of why these three should co-occur, the field has few grounds to expect they will, and less basis to explain when a study finds two of the three without the third. This review develops that theoretical account (Section 2.4) and uses it to interpret what the corpus does, and does not, show.

A geographic blind spot compounds the problem. Most work comes from Asia and Europe. Latin American universities, which contend with constrained budgets, politicized governance, and deteriorating infrastructure, are barely visible in the literature (Scavarda et al., 2023; Tello and Aguilar, 2024). Ecuador, whose constitution enshrines buen vivir and the rights of nature, presents a case the field has largely overlooked (Álava-Atiencie and Quinde-Lituma, 2023).

This review asks: how does the HEI literature integrate transformational leadership, sustainability governance, and organizational maturity in relation to sustainable energy performance, and what gaps persist? Four sub-questions guide the analysis: (SQ1) Which TL dimensions appear in HEI sustainability work? (SQ2) What governance arrangements facilitate campus energy transition? (SQ3) What maturity models have been proposed for HEI sustainability? (SQ4) How far does the literature go in combining the three constructs, and where does Latin America fit?

This review makes three contributions. Theoretically, it grounds the relationship between transformational leadership, sustainability governance, and organizational maturity in a mechanistic account rather than a descriptive juxtaposition, using dynamic-capabilities and organizational-learning theory to explain why the three constructs should be expected to co-occur (Section 2.4). Empirically, it provides the first bibliometric and thematic map of how these three constructs are covered, individually and jointly, in the HEI sustainability literature, documenting construct-coverage asymmetries with maturity as the weakest link. Conceptually, it proposes a Governance and Maturity Model (MGM) and a five-proposition research agenda aimed at scaling solar energy and efficiency projects on university campuses, with particular attention to Latin American contexts where structured maturity diagnostics appear most needed.

Conceptual background

Transformational leadership in HEI sustainability

Bass's (1985) model describes leaders who motivate through idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration. In sustainability research, this has been relabelled “green transformational leadership” (GTL), referring to leaders who promote environmental awareness and green innovation (Bahzar, 2019; Farooq et al., 2024; Mukhtar et al., 2025). Within universities, TL has been connected to sustainability-oriented teaching (Eustachio et al., 2024), citizenship behavior for CSR (Alshihabat and Atan, 2020; Mansaray and Atan, 2025), strategic planning (Habeeb and Eyupoglu, 2024), and innovative work behavior (Connie et al., 2025). Recent work also links digital leadership capabilities to green innovation via knowledge sharing (Shen et al., 2025).

A consistent limitation across these studies is that TL is treated as an independent variable predicting individual-level outcomes. How leadership interacts with governance structures or with the institution's maturity level to produce systemic change receives little attention (Liao, 2022). Sustainability transitions need more than motivated people. They need system-level capacity that outlasts any individual leader.

Sustainability governance in higher education

Sustainability governance in HEIs covers the structures through which universities organize their sustainability efforts (Bauer et al., 2018). The most systematic work comes from a German research group that built the “governance equalizer,” a tool mapping five dimensions: politics, profession, organization, knowledge, and public discourse (Bauer et al., 2018; Niedlich et al., 2020a). They applied it to identify governance cultures across German universities (Bauer et al., 2020) and developed maturity scales for each dimension (Niedlich et al., 2020b).

Other researchers have examined shared governance in university athletics (Barrett et al., 2019), governance networks in African HEIs (Jimu and Rennkamp, 2024), web-based quality frameworks (Salvioni et al., 2017), HEInnovate self-assessment (Dieguez, 2023), and ESG governance in Saudi Arabia (Alhusaini, 2025).

A recurring finding across these studies: governance on its own is not enough. Without leadership drive and institutional capacity, governance structures remain on paper (Robinson and Laycock Pedersen, 2021; Roos et al., 2022).

Organizational maturity and diagnostic frameworks

Maturity models originate from the Capability Maturity Model (Humphrey, 1989), which rates organizational capability on a scale from ad hoc to optimized. In HEI sustainability, maturity has surfaced through the governance equalizer's maturity scales (Niedlich et al., 2020b), leadership maturity for Six Sigma (Anthony and Antony, 2021), balanced scorecard adaptations (Engin, 2022) and bibliometric reviews of sustainable performance trends (Umar et al., 2024).

Beyond these few studies, coverage is sparse. Most studies mention maturity in passing: as a contextual variable, a limitation, or a direction for future work. No standalone maturity model for assessing a university's capacity to manage and scale energy sustainability projects was identified in the reviewed corpus. That gap is what prompted this review.

A theoretical lens on the three constructs

The three constructs are not merely co-present; they are linked mechanisms in a single capability-building process, and reading them separately is what has kept the field from seeing how they depend on one another. Dynamic-capabilities theory offers the connecting mechanism: an organization's capacity to adapt rests on three sequential micro-foundations—sensing opportunities and threats, seizing them through decisive resource allocation, and reconfiguring (or transforming) the organization's asset base so the change sticks (Teece, 2007). Mapped onto the three constructs, transformational leadership performs the sensing function: idealized influence and inspirational motivation direct an institution's attention toward an emerging opportunity or threat, such as an unmet SDG 7 target or a funding window for solar retrofits, in a way that routine management does not (Bass, 1985; Umar et al., 2025; Farooq et al., 2024). Sustainability governance performs the seizing function: once an opportunity has been sensed, it is the governance structure, not the leader alone, that allocates decision rights, mobilizes stakeholders, and commits resources to act on it (Bauer et al., 2018; Niedlich et al., 2020a). Institutional theory explains why that seizing is not automatic: universities adopt governance structures partly in response to coercive, normative, and mimetic pressures, and whether the resulting structure is substantive or symbolic depends on which of these pressures dominates (Bagire et al., 2024; Lo, 2015). Organizational maturity, finally, performs the reconfiguring function: it is the accumulated, staged evidence that sensing and seizing have actually changed what the institution can do, not just what it says it will do (Humphrey, 1989).

Organizational-learning theory supplies the bridge the field has left unbuilt between leadership and maturity directly, without governance as an intermediary. A single episode of sensing and seizing changes a project; it becomes maturity only when the institution learns from it in a way that revises its underlying routines rather than just its outputs, the distinction organizational-learning theory draws between single-loop learning (correcting an action) and double-loop learning (revising the assumptions and structures that produced it) (Argyris and Schön, 1978). Transformational leadership is what tips an institution from single-loop toward double-loop learning: intellectual stimulation and individualized consideration are, in organizational-learning terms, mechanisms for questioning existing assumptions rather than merely optimizing existing procedures (Farooq et al., 2024; Bass, 1985). Repeated cycles of double-loop learning, sustained across personnel changes, are what a maturity stage actually consists of. This is a direct leadership-to-maturity channel that does not require governance as a mediator, and its near-total absence from the reviewed corpus (Section 5.2) is therefore a gap in the literature's mechanisms, not evidence that the channel does not exist.

Read together, the three theories describe a cycle rather than a line: leadership senses, governance seizes, maturity reconfigures and accumulates capability, and that accumulated capability changes what the institution is subsequently able to sense and seize, feeding back into the leadership function at a higher baseline. This cyclical, feedback-bearing structure, developed further and diagrammed in Section 5.7, is what distinguishes the account offered here from a normative claim that leadership, governance, and maturity would simply be nice to combine; it specifies the mechanism by which each construct produces the conditions for the next, and motivates both the coding scheme and the research agenda below.

The integration gap

The case for combining TL, governance and maturity is simple. Leadership provides direction and energy. Governance provides institutional structure. Maturity provides a staged diagnostic so the institution knows where it stands and what to build next. Drop any one and things break: leadership without governance does not survive personnel changes; governance without maturity produces policies the institution cannot actually carry out; maturity without leadership diagnoses problems but cannot drive change.

A few studies have connected two of the three. Umar et al. (2025) combined change management with TL and knowledge management. (Robinson and Laycock Pedersen 2021) paired governance with resilience. Shen et al. (2025) linked digital leadership to green innovation. Alhazemi (2025) integrated leadership, culture and policy. But, we did not identify work that examined all three together. This review checks where the connections have been made within the corpus and where they have not.

Protocol and reporting

This review was conducted and reported following the PRISMAL 2020 statement (Page et al., 2021). An a priori protocol, named SIGMA (Systematic Integrated Governance–Maturity Analysis), was written before the searches were run. It specified the research question, the PICO framing, the eligibility criteria, the full search strings, and the extraction and synthesis plan. In PICOS terms, the review targets Populations: higher education institutions (public and private universities and campuses); Intervention/phenomena of interest: transformational leadership, sustainability governance, and organizational maturity applied to campus energy and sustainability management; Comparators: none required, as the review maps construct coverage rather than comparing interventions; Outcomes: sustainable energy performance and related organizational outcomes (efficiency, project scaling, and institutionalization); and Study designs: peer-reviewed journal articles, reviews, and indexed conference papers. The protocol was not lodged in a public registry; it is provided in full as Supplementary material so that the procedure can be audited and reproduced.

Eligibility criteria

Records were included when they were (I1) a peer-reviewed journal article, review, or indexed conference paper; (I2) published between January 2014 and March 2026; (I3) written in English, Spanish, or Portuguese; (I4) addressing at least one of the three focal constructs, namely transformational leadership (TL), sustainability governance, or organizational maturity, within a higher education institution (HEI) or an institutional energy-management setting; (I5) available in full text; and (I6) carrying a verifiable DOI or formal index record.

Records were excluded when they (E1) treated energy only at the residential, industrial, or grid level with no link to HEIs or institutional governance; (E2) were ineligible document types (books, book chapters, conference reviews, editorials, errata, notes, letters, retractions, theses, internal reports, or news items); (E3) fell outside the temporal window; (E4) modeled purely technical systems (panels, batteries, networks) with no organizational, governance, or behavioral component; (E5) had no accessible full text; or (E6) reported no identifiable methodology.

The 2014–2026 window was chosen deliberately. It opens the year before the 2015 adoption of the Sustainable Development Goals and the Paris Agreement, which reframed institutional energy action, and it coincides with the consolidation of the HEI sustainability-governance literature, including the German “governance equalizer” programme (Bauer et al., 2018). Foundational conceptual sources predating the window (e.g., Bass, 1985; Humphrey, 1989) were retained as theoretical anchors but were not treated as corpus records.

Information sources and search strategy

The formal, traceable search was executed in two indexed databases: Scopus (Elsevier) as the primary source and Web of Science (Clarivate, Core Collection) as the complementary source for cross-database validation. No other platform contributed records to the corpus. Full-text retrieval and regional context were supported by publisher platforms and Google Scholar, but these were not used as sources of identification and none of their items entered the corpus. AI-assisted discovery tools were not used to identify, screen, or select records.

Identification rested on a single integrated master string, S_MASTER, that required the simultaneous presence of (a) at least one construct term, (b) at least one HEI term, and (c) at least one energy/sustainability term. In Scopus the string was run in the TITLE-ABS-KEY field; in Web of Science it was adapted to the TS= (Topic) field. Asterisk truncation captured morphological variants. The exact query was:

TITLE-ABS-KEY ((“transformational leadership” OR “organizational maturity” OR “process maturity” OR “capability maturity” OR “energy management maturity” OR “sustainability governance” OR “stakeholder governance”) AND (“higher education” OR university OR campus OR “higher education institution*” OR HEI) AND (“energy management” OR “energy efficiency” OR sustainability OR “renewable energy” OR “green campus” OR “net zero” OR “energy transition”))

Executed in March 2026, S_MASTER returned 127 records in Scopus and, in its TS= adaptation, 103 records in Web of Science, for 230 records identified. Six construct-block strings (S1–S6), each targeting one thematic axis in isolation, were used during protocol development to calibrate terminology and to confirm the coverage of the master string; because each block returned between roughly 9.5 × 10 2 and 1.7 × 10 4 records for a single axis, none was used to compile the corpus. The six block strings, their field settings, and their individual counts are reported in full in Supplementary Appendix A, so that the derivation of S_MASTER is fully reproducible.

Selection process

Selection followed the four PRISMAL phases (Figure 1). Of the 230 records identified (Scopus n = 127, WoS n = 103), 64 inter-database duplicates were removed after DOI-level verification, and 37 ineligible document types indexed in Scopus (21 book chapters, 7 books, 4 conference reviews, 3 errata, 1 note, and 1 letter) were removed before screening, leaving 129 unique records (90 from Scopus, 39 from WoS).

At title-and-abstract screening, 56 records were excluded: 22 for falling outside the HEI or institutional-governance context (E1), 33 for being purely technical with no organizational component (E4), and 1 residual duplicate (E3). Seventy-three records advanced to full-text assessment. At the full-text stage, 39 were excluded with a recorded reason: 17 because no focal construct was central or the setting was not an HEI (E4), 20 because the full text could not be obtained (E5), and 2 because no methodology was reported (E6). Thirty-four studies met all criteria and constitute the final corpus.

Screening and full-text assessment were carried out independently by two reviewers. Each reviewer applied the eligibility criteria to every record; disagreements were logged and resolved by consensus discussion, with the a priori criteria as the reference. A record was retained only when both reviewers agreed it was eligible. Records excluded at full text, with their individual exclusion codes, are listed in Supplementary material to allow independent audit.

Data extraction and construct coding

For each included study, two reviewers independently extracted: authors, year, country or region, study design, method, focal construct(s), coverage of TL, governance, and maturity, and thematic category. Extractions were compared and reconciled by consensus.

Construct coverage was coded on a three-level scale (Yes, Partial, or No) using an a priori rubric with explicit operational definitions (Table 1). To limit subjectivity, “integration” was defined narrowly and transparently: a construct was coded Yes only when it was a central variable, that is, when it was theorized, measured, or analyzed as such rather than mentioned in passing. This operational threshold is what the review's coverage claims refer to; studies that engage a construct under alternative terminology at a peripheral level are captured by the Partial code, not lost. The two reviewers coded all 34 studies independently; agreement was high, and the residual disagreements were resolved by consensus against the rubric. Because coding was performed to consensus rather than reported as a single-rater judgment, the construct-coverage results rest on agreed classifications rather than on one reviewer's discretion. Inter-rater reliability was assessed prior to consensus discussion using Cohen's Kappa on the full 34-study corpus. Agreement was almost perfect for transformational leadership (κ = 0.945) and sustainability governance (κ = 1.000), and substantial for organizational maturity (κ = 0.796) (Landis and Koch, 1977). Discrepancies were resolved through discussion against the pre-specified coding rubric (Table 1).

Thematic synthesis followed Thomas and Harden (2008): open coding of each study's contribution, then axial coding into six categories (C1–C6). Bibliometric indicators were treated descriptively (temporal, geographic, methodological, and construct-coverage distributions) and are reported as a descriptive map of the field rather than as a formal bibliometric network analysis, which the size of the corpus (n = 34) would not support robustly.

Quality assessment

The methodological quality of each included study was appraised using the Mixed Methods Appraisal Tool (MMAT), version 2018 (Hong et al., 2018), a validated instrument designed for reviews that include heterogeneous study designs. MMAT was selected over JBI or CASP because the corpus combines cross-sectional survey work modeled with PLS-SEM, qualitative multi-case designs, and conceptual contributions; the design-specific alternatives would have required several separate instruments whose ratings are not directly comparable.

Each study was first screened against the two MMAT screening questions, then assigned to one of five methodological categories and rated on the five items specific to that category. Ratings were recorded as yes, no, or cannot tell. Both reviewers applied the instrument independently and reconciled discrepancies by consensus against the full text. Unlike the construct coding described above, the pre-consensus ratings for the quality appraisal were not retained as a separate record, so inter-rater agreement is not reported for this instrument; the appraisal reflects the reconciled, consensus-based rating for each item. Following the explicit guidance of the MMAT developers, no overall numerical quality score was computed, since such scores obscure which specific methodological dimension is weak; ratings are reported item by item in Supplementary Appendix D.

Studies that made a conceptual or theoretical contribution without collecting empirical data fall outside the scope of MMAT and were recorded as not appraisable rather than rated as low quality. Their contribution to the synthesis is thematic rather than evidential, and no coverage finding rests on them alone.

The four criteria adapted from Petticrew and Roberts (2006) that guided the first round of appraisal were retained as a complementary layer, since they capture reporting transparency across designs in a way the design-specific MMAT items do not. Studies flagged as weak under either layer were retained given the small corpus, and their influence on the synthesis is discussed where relevant.

Risk of bias across studies

The principal review-level risk of bias is database coverage. Restricting identification to Scopus and Web of Science may have missed work indexed elsewhere, including gray literature and institutional sustainability reports. Language restriction to English, Spanish, and Portuguese introduces further selection bias against relevant research published in German or Mandarin. Because organizational maturity appears under heterogeneous labels across disciplines, the search may also under-detect maturity-related work framed in alternative terminology. For these reasons, coverage findings are reported throughout as properties of the reviewed corpus (“not identified within the reviewed corpus”), and are not presented as evidence that a phenomenon is absent from the wider literature.

Temporal distribution

The 34 studies span 2014–2026. The early period (2014–2018, n = 5, 14.7%) clusters around the German governance group and the first campus case studies. Output rose in 2019–2022 (n = 9, 26.5%), as PLS-SEM studies of leadership spread across Asia. The most recent period (2023–2026, n = 20, 58.8%) accounts for well over half the corpus (Figure 2). The literature is therefore recent and still forming.

Geographic distribution

Asia and the Middle East produced 19 studies (55.9%), led by Indonesia (4), Malaysia (4), Saudi Arabia (3), Oman (2), and Turkey (2). Europe contributed 5 (14.7%), with Germany accounting for 3, all from the Bauer–Niedlich governance group. Africa produced 4 (11.8%: Nigeria, South Africa, Uganda, and Sierra Leone). Latin America produced 2 (5.9%: Colombia and Ecuador, one each), North America 2 (5.9%), and two studies (5.9%) were multi-country in scope (Figure 3). Of the 34 studies, 20 were indexed in both Scopus and Web of Science and 14 in Scopus only, indicating substantial cross-database overlap for the master query.

The geographic distribution mirrors a methodological split. Asian studies favor quantitative designs, particularly PLS-SEM; European studies are predominantly qualitative and conceptual; the two Latin American studies are exploratory in design. An integrative model would benefit from mixing these approaches.

Construct coverage

Leadership has the widest direct coverage: 17 studies (50.0%) treat it as a central variable, rising to 20 (58.8%) when partial coverage is included. Governance follows with 15 direct (44.1%) and 24 (70.6%) including partial coverage. Maturity is far behind: only one study (2.9%) addresses it as a central variable, the balanced-scorecard application of Engin (2022), though six studies (17.6%) mention it partially. Maturity surfaces in discussion but is almost never placed at the center of an analysis (Figure 4). This asymmetry is a coverage property of the reviewed corpus and should not be read as evidence that maturity research is absent from the wider literature, where it appears under heterogeneous labels.

Co-occurrence is consistent across both coding thresholds. Counting direct and partial coverage together (Figure 5), 10 studies (29.4%) pair leadership with governance, and 6 (17.6%) pair governance with maturity, but no study combines leadership with maturity in the absence of governance, and no study reaches all three. Under strict measurement, counting direct coverage only (Figure 6), the corpus is markedly unidimensional: 15 studies (44.1%) address only leadership, 13 (38.2%) only governance, and 1 (2.9%) only maturity; 3 studies (8.8%) have no construct as a central variable. Two studies (5.9%) treat both leadership and governance as central, but the triple combination and the leadership–maturity pairing are not identified under either threshold. Across the corpus, 85.3% is built around a single central construct.

Thematic synthesis

Six categories emerged from the coding (Figure 7): C1, assessment frameworks and maturity (n = 2, 5.9%); C2, TL and sustainable performance (n = 13, 38.2%); C3, sustainability governance in HEIs (n = 10, 29.4%); C4, campus energy and environmental management (n = 4, 11.8%); C5, convergence of leadership, governance and maturity (n = 2, 5.9%); and C6, Latin American and peripheral contexts (n = 2, 5.9%). One further study (2.9%) sits at the C2–C3 boundary, combining leadership and governance. The distribution confirms the coverage pattern: leadership-centered and governance-centered work dominates, while maturity-centered and integrative work is thin.

Energy performance as an outcome

Because the review's framing question concerns sustainable energy performance, we examined how directly the corpus engages energy outcomes as opposed to sustainability in general. Only the four C4 studies plus a small number of others (Bahzar, 2019; Saleh et al., 2015; Bagire et al., 2024) treat campus energy management, efficiency, or solar deployment as an explicit outcome; the majority address sustainability governance or leadership in broad terms, with energy implicit. This is itself a finding: within the reviewed corpus, the link between the three organizational constructs and measurable energy performance is rarely examined directly. This is the gap that the proposed model addresses.

Evolution of the field

A plausible expectation is that the corpus reflects an ordered progression from leadership, to governance, to maturity, as the field matured. Tabulating direct construct coverage by publication period shows the opposite trajectory. In 2014–2018, governance was central in 60.0% of studies and leadership in none; by 2019–2022, leadership had risen to 33.3% while governance remained dominant at 66.7%; by 2023–2026, leadership was central in 70.0% of studies while governance had fallen to 30.0% (Table 2). The field's center of gravity moved from governance to leadership, not from leadership toward maturity. Maturity itself never established a foothold: it was central in exactly one study across the entire window, Engin (2022) in 2022, an isolated case rather than an emerging third phase.

This reading is consequential for the model proposed below. Maturity is not the phase the field is already moving toward as governance work matures into something more diagnostic; it is a phase that, within this corpus, did not occur at all. The shift from governance to leadership as the dominant focus suggests scholarly attention chases the most recently prominent construct rather than building cumulatively toward institutional capability assessment, which is precisely the gap the MGM is designed to fill rather than a gap it can assume the field is already closing on its own.

The maturity gap

The clearest pattern is the distance between governance coverage and maturity coverage: 44.1% vs. 2.9%. Governance has a developed body of work, particularly from the German cluster. Maturity does not. It appears in discussion sections and future-research paragraphs but is rarely operationalized with instruments, staged typologies, or diagnostic protocols. Of the studies that engage maturity at all, only Engin (2022) places it at the center, and there through a balanced-scorecard lens rather than a sustainability-specific maturity model; the remainder (Niedlich et al., 2020b; Štrukelj et al., 2023; Robinson and Laycock Pedersen, 2021) treat it as one dimension among several. Within this corpus, no standalone maturity model for HEI energy sustainability was identified.

In practical terms, a university that wants to scale solar installations or retrofit buildings for efficiency has no validated way to locate its readiness on a continuum. Maturity, as a construct, has a clear structure to offer: rooted in the Capability Maturity Model (Humphrey, 1989), it stages capability from ad hoc, through repeatable and defined, to managed and optimizing, and it can be operationalized across dimensions such as strategy, governance, technical systems, data, and culture. The governance equalizer's maturity scales (Niedlich et al., 2020b) are the closest existing tool, but they are embedded in a governance framework rather than designed as a standalone energy diagnostic. Global South universities are affected most, because they typically start from a lower capability baseline and need the sharpest possible picture of what to build first.

Under strict measurement, the concentration is more pronounced: 85.3% of the corpus is organized around a single central construct, and no study treats all three as central variables (Figure 6). We frame this as a property of the reviewed corpus rather than proof of absence in the wider literature; the search was limited to two databases and three languages, and maturity in particular travels under many labels.

Read through the theoretical lens of Section 2.4, the maturity gap is a gap in the reconfiguring stage of the sensing-seizing-reconfiguring cycle. The field has produced considerable work on sensing (leadership) and seizing (governance), but comparatively little on the accumulated, staged evidence that those first two stages actually changed institutional capability. A plausible reason is methodological: sensing and seizing are more tractable to study cross-sectionally with a single survey wave, while reconfiguring, by definition, requires observing an institution over time as capability accumulates, a design few studies in this corpus attempt (Section 5.6). Institutional theory adds a second, complementary reason: coercive and mimetic pressures push universities to adopt visible governance structures quickly, because those structures are what external accreditors and rankings observe, while maturity is an internal, harder-to-signal property that generates less immediate legitimacy payoff for the institution that invests in measuring it.

Terminological bias and the maturity signal

The near-absence of maturity from the corpus raises a prior question: does it reflect a genuine scarcity in the literature, or an artifact of how S_MASTER was constructed? The distinction matters, because the review's central claim rests on maturity being the least-developed of the three constructs. Two features of the master string bound the inference. First, S_MASTER required the simultaneous presence of a construct term, an HEI term, and an energy or sustainability term; maturity work conducted in industrial, corporate, or public-sector settings, or framed as process improvement without an explicit higher-education or energy anchor, falls outside the query by design. Second, the maturity vocabulary in S_MASTER was deliberately narrow: it covered “organizational maturity,” “process maturity,” “capability maturity,” and “energy management maturity,” but did not include several labels under which the same construct travels, such as CMM, CMMI, EMMM, ISO 50001 readiness, or the generic “maturity model,” all of which appeared in the wider scoping string S3 (Supplementary Appendix A) but were not carried into the integrated query. Maturity research indexed only under those alternative terms would therefore be under-detected. We read the 2.9% figure accordingly: it is a reliable statement about how often maturity is treated as a central construct within the reviewed corpus, not a census of the wider literature, in which maturity is demonstrably studied under heterogeneous labels (Humphrey, 1989; Niedlich et al., 2020b). This is why every coverage claim in this review is bounded to the corpus, and why the agenda treats the maturity gap as a hypothesis to be tested against broader term sets rather than as an established void.

The absent bridge between leadership and maturity

No study in the corpus paired transformational leadership with maturity while leaving governance aside. Leadership scholars study vision, motivation, and follower behavior; maturity scholars study process levels and organizational capability. Within this corpus the two literatures run in parallel and meet only when governance sits between them.

Leadership maturity models do exist in professional practice. The Leadership Maturity Model (Anderson, 2024) outlines developmental levels, and the Leadership Maturity Framework (Cook-Greuter, 2013) maps stages of ego development; both are practitioner instruments rather than peer-reviewed empirical studies, and we cite them as such. Within the peer-reviewed HEI sustainability corpus reviewed here, none of these has been tested, and no study examines how transformational leadership builds organizational maturity without governance as a mediator. The concept lives largely in practitioner settings and was not identified in the academic HEI-sustainability work captured by our search.

For model building this pattern matters, but it should be read as a research direction rather than a proven void. Section 2.4 specifies why a direct channel should exist theoretically: organizational-learning theory predicts that transformational leadership drives double-loop learning directly, independent of governance mediation, and repeated double-loop learning cycles are what a maturity stage consists of Argyris and Schön (1978) and Farooq et al. (2024). Which specific leadership behaviors move an institution from ad hoc sustainability efforts to systematic ones, and whether that movement genuinely bypasses governance or simply passes through an unmeasured governance variable, is an open empirical question this corpus cannot resolve. That this pairing was not identified may also reflect the search terminology; future reviews with broader term sets and databases should test the boundary directly.

Why governance moved faster than maturity, and why leadership overtook governance

Three patterns from Section 4 deserve a mechanistic explanation rather than a restatement. First, why is maturity still rarely studied? Under dynamic-capabilities theory, sensing and seizing are episodic and observable within a single study; reconfiguring is cumulative and only becomes visible over multiple cycles, so it is simply harder to publish a cross-sectional maturity study with a compelling result. Second, why did governance develop faster than maturity, particularly in the German cluster? Institutional theory suggests governance responds to external legitimacy pressures (rankings, accreditation, and funder expectations) that reward visible structures on a shorter timescale than internal capability-building does, so governance research had an earlier and steadier stream of empirical variation to study. Third, why is transformational leadership studied more at the individual level than at the organizational level, as Section 2.1 noted? The sensing function that TL performs in the theoretical model is, by construction, an individual cognitive and motivational act; the corpus's near-universal use of individual-level TL scales is not a methodological oversight so much as a direct consequence of what sensing is. What the corpus has not yet done is trace how an individual leader's sensing act propagates into the governance structures that seize it and the maturity stages that reconfigure around it, which is precisely the systemic linkage the MGM proposes to test.

Why governance is rarely tied to physical energy outcomes

Section 4.5 reported that only the four C4 studies, together with a small number of others (Bahzar, 2019; Saleh et al., 2015; Bagire et al., 2024), treat measurable energy performance as an explicit outcome; the majority address governance or leadership in general sustainability terms. That most of the corpus stops short of linking organizational structures to physical consumption data (metered electricity, energy-use intensity, avoided kWh, or verified solar output) deserves explanation rather than restatement. Three mechanisms are plausible. First, a disciplinary and data divide: governance and leadership are studied predominantly by management and education scholars using perceptual, survey-based instruments, whereas physical energy data reside in facilities and engineering systems that those researchers rarely access or co-author with, so the two evidence bases seldom meet in a single study. Second, a measurement-horizon mismatch: the effect of a governance structure on physical consumption is lagged and confounded by weather, occupancy, and unrelated retrofits, which makes it far harder to attribute cleanly than a cross-sectional association between leadership scores and self-reported behavior. Third, read through the lens of Section 2.4, the corpus concentrates on the sensing and seizing moments of the capability cycle (both legible through perceptual measures), while physical energy performance is the downstream outcome of the reconfiguring stage, observable only once capability has accumulated and been exercised on plant and buildings. The scarcity of energy-outcome studies is thus the maturity gap seen from the outcome side: the field has measured the antecedents of energy performance far more often than energy performance itself. This asymmetry is what the title's “toward” signals, and the one the MGM is designed to close by making the maturity-to-energy-performance link an explicit, measurable stage rather than an assumed endpoint.

From evidence to agenda: guarding against confirmation bias

Because this review proposes a specific model, we have been deliberate in separating what the corpus shows from what we propose. The coverage asymmetries, the unidimensional majority, and the empty leadership–maturity cell are empirical patterns in the coded data; the MGM is a response to those patterns, not a premise imposed on them. Where the evidence is thinnest, on maturity, our claims are correspondingly cautious and framed as corpus-level absence. The research agenda that follows is therefore presented as hypotheses to be tested, not as conclusions the review has established.

The Latin American gap

Two studies out of 34, one of them from Ecuador. This is a thinner Latin American presence than broad sustainability-in-HEI reviews report, and it is consequential. Latin American universities operate under conditions in which structured maturity diagnostics would be particularly useful: constrained budgets force prioritization, politically exposed governance creates instability across administrative cycles, and infrastructure deficits demand staged investment planning (Scavarda et al., 2023). A descriptive count of studies understates the problem; the deeper issue is that the region's specific governance conditions (short administrative horizons, dependence on public financing, uneven data systems) are exactly the conditions a maturity model is designed to diagnose, yet no corpus study does so.

Ecuador is an instructive case. Its single corpus study (Álava-Atiencie and Quinde-Lituma, 2023) examines transformational leadership and the social-entrepreneurial profile at the University of Cuenca but says nothing about governance structures or maturity for energy management. The country's constitutional commitments to buen vivir and the rights of nature create a normative environment that could either enable or constrain sustainability governance in universities, and that intersection is unexplored. A context-sensitive MGM designed for Ecuadorian universities would be novel, and its logic could transfer to comparable Andean and Latin American settings.

Methodological patterns

Quantitative modeling dominates the corpus: roughly 10 studies use PLS-SEM or covariance-based SEM, typically on single-country samples, a reasonable choice for testing multi-variable models with modest samples (Alhazemi, 2025; Eustachio et al., 2024; Farooq et al., 2024; Alshihabat and Atan, 2020). Yet, the qualitative governance studies, particularly the German cluster's multi-case designs with expert interviews (Niedlich et al., 2020a) and participatory work on campus governance (Zen et al., 2016), have produced the frameworks with the most practical traction. Mixed-methods designs that pair quantitative maturity assessment with qualitative governance analysis would go further than either alone. The near-absence of longitudinal designs is a further open frontier: without tracking how institutions move through maturity stages over time, the field cannot separate correlation from mechanism.

Toward an integrative model

These patterns inform the model in Figure 8. Earlier drafts of this model, in common with much of the literature it synthesizes, framed the relationship as a one-directional sequence: leadership drives governance, governance builds maturity, maturity produces energy performance. Sections 2.4 and 5.1–5.3 show why that linear reading understates the mechanism: because organizational maturity is itself the accumulated capability that changes what an institution can subsequently sense and seize, the model is better specified as a cycle with an explicit feedback loop rather than a one-way chain. Figure 8 therefore depicts transformational leadership as the micro-level driver that performs the sensing function, sustainability governance as the meso-level structure that performs the seizing function under institutional pressure, and organizational maturity as the developmental spine that performs the reconfiguring function, tracking capability from ad hoc to optimizing, with sustainable energy performance as the outcome each cycle is meant to produce. The feedback arrow closes the loop: as maturity accumulates, it raises the baseline capability from which the next round of leadership sensing and governance seizing operates, so that a mature institution is not just better at delivering energy projects but faster at recognizing and mobilizing around the next one. The arrows throughout are propositions, not established effects; the next section states them as such.

This review examined 34 studies from Scopus and Web of Science (2014–2026), mapping where transformational leadership, sustainability governance, and organizational maturity meet in the higher education literature. Three findings are central here, each a property of the reviewed corpus.

Maturity is the construct the field has engaged least. Governance (44.1% direct coverage) and leadership (50.0%) have established research traditions with validated tools; maturity was central in a single study and was never identified paired with leadership alone. Under strict measurement, no study treated all three as central variables. Within this corpus, universities lack diagnostic instruments to assess their own readiness for sustainability transitions.

The field is growing quickly, with 58.8% of the corpus appearing between 2023 and 2026, but unevenly, and its center of gravity moved from governance toward leadership rather than from leadership toward maturity (Section 4.6). Quantitative work dominates without the contextual depth that qualitative designs provide, and cross-sectional data cannot reveal how institutions develop over time. Foundational integrative frameworks are still being built.

Latin America accounts for 5.9% of the corpus, with a single Ecuadorian study. The region's universities need governance and maturity approaches fitted to their realities, and Ecuador's normative framework around buen vivir and rights of nature could enable or constrain sustainability governance in HEIs, which remains an open question.

Beyond mapping these patterns, this review's central contribution is theoretical: it is the first, to our knowledge, to specify a mechanism, grounded in dynamic-capabilities and organizational-learning theory, by which transformational leadership, sustainability governance, and organizational maturity are expected to co-occur, rather than simply documenting that they often do not. That mechanism reframes the field's best-known finding, the leadership-maturity gap, from a fact about the literature into a testable claim about an under-examined causal channel (Proposition 2). It extends the literature on HEI sustainability governance, which has so far treated governance largely as an endpoint, by repositioning it as one stage in a cyclical capability-building process with organizational maturity as the accumulating, feedback-bearing core. The five propositions in the research agenda operationalize that mechanism as testable hypotheses rather than settled conclusions, and the Governance and Maturity Model (MGM) is offered as the conceptual structure through which future empirical work, beginning with Latin American HEIs, can test it.

The review has limitations that bound these claims. Two databases, three languages, and a 2014–2026 window may have excluded relevant work, including gray literature and reports indexed elsewhere; because organizational maturity appears under varied terminology, the search may under-detect it specifically. Coverage findings are therefore reported as “not identified within the reviewed corpus” rather than as absolute absence. The three-level coding, though performed independently by two reviewers with almost-perfect to substantial inter-rater agreement (κ = 0.796–1.000) and reconciled by consensus, still involves judgment at the Partial boundary. Future reviews should widen the database and language set, add gray literature from institutional sustainability reports, and test whether alternative search terms surface work this review did not capture, particularly on organizational maturity, where the corpus signal is weakest and the stakes for the argument are highest.