2.1 Theoretical Background

A stream of prior studies has utilised the upper echelons theory (UET) to explain the influence of top executives on environmental activities (e.g., Marie et al. 2025; Bilal et al. 2023; Bilal Khan et al. 2024; Huang et al. 2023) and environmental reporting (Shahab et al. 2020). According to UET (Hambrick 2007; Hambrick and Mason 1984), top executives possess distinct observable attributes such as age, personality, career experience and education which significantly shape organisational outcomes and performance (Ha, Kang, and Kwon 2024; Zhao and Qu 2023). UET suggests that scholars can use observable managerial characteristics as indicators (Hambrick and Mason 1984) such as age, tenure and experience to explore how executives personal attributes manifest in strategic decisions (Ullah, Jiang, and Elamer 2024; Wang et al. 2016). The literature has continued to stress that the unique characteristics of top executives influence strategies and outcomes, and their personal attributes impact their decision-making abilities (Carvajal, Nadeem, and Zaman 2022). The theory posits that as VCs are spotlighted like CEOs, their characteristics may influence sustainability strategies, warranting an investigation into the interplay between VC horizon and sustainability performance (Bugeja et al. 2021).

So far, the literature emphasises the importance of personal attributes in shaping sustainability strategies (Albitar, Abdoush, and Hussainey 2023; Andersson et al. 2022; Tang et al. 2023). Despite clear evidence that the personal characteristics of the CEO matters in the corporate context (Al-Asmakh, Elamer, and Uadiale 2024; Usman, Nwachukwu, and Ezeani 2022), recent studies have extended this focus to examine the role of the VCs characteristics in HEIs, revealing that VC leadership is a key predictor of disclosures (e.g., Cheah et al. 2023). For instance, Elmagrhi et al. (2021) find that VC tenure and board specific determinants of committee presence and board meetings positively influence governance disclosures in UK HEIs, while factors like VC age and gender show negative associations. Lucey, Urquhart, and Zhang (2022) observe that UK VCs are not over-compensated, with female VCs earning less than their male counterparts, and longer-tenured VCs receiving higher salaries. Similarly, Elmagrhi and Ntim (2022) reveal a negative relationship between VC pay and HEIs' long-term social performance but a positive relationship with short-term performance. More recently, Cheah et al. (2023) highlight VC power's effect on HEI efficiency.

Building upon the existing body of research, it is evident that the personal characteristics of VCs play a critical role in shaping various aspects of HEIs, particularly in the area of sustainability performance. The UET provides a valuable framework for understanding how VCs' attributes affect their strategic decisions and, consequently, the institution's sustainability performance. So far, the theory has been instrumental in highlighting how these characteristics influence leadership decisions and there is growing recognition of its relevance to the HEI context (Elmagrhi and Ntim 2023). However, while these studies underscore the importance of VC characteristics in shaping governance and financial outcomes, they also point to a gap in understanding how these personal attributes affect sustainability performance. The interaction between a VC's personal horizon—encompassing their long-term vision and commitment to sustainability—and the institution's sustainability strategies remains underexplored. Investigating this relationship through UET could reveal how leadership characteristics directly influence institutional sustainability performance, offering deeper insights into the link between executive attributes and broader sustainability outcomes in HEIs.

2.2 Hypotheses Development

3.1 Data

To examine these hypotheses, we have utilised a hand-collected dataset for our research, incorporating relevant data gathered from the Times Higher Education Ranking, Complete University Guide websites, universities' annual reports and their official websites. As most ranking league tables have only recently begun to focus on sustainability-related information, our sample includes all universities in the United Kingdom from 2018 to 2022. ¹ Additionally, using the Higher Education Statistics Agency (HESA) website, we identified the entire population of HEIs—including universities, colleges and other HEIs—in the United Kingdom as of 31 December 2022. We then visited the websites of all 164 identified universities to download their annual reports for the years 2018–2022. After excluding missing values across all variables, our final sample consists of 523 firm-year observations.

3.2 Variables

Our dependent variable, ‘sustainability’ is constructed using a machine-learning approach for textual analysis, using Python software and an array of essential libraries (Bochkay et al. 2023; Ibrahim, Elamer, and Ezaet 2021). The incorporation of Python in our research for conducting textual analysis confers notable advantages in comparison to prior methods. Python as a programming language facilitates the integration of diverse NLP libraries and tools, thereby streamlining the entire analysis process (Bhandari, Ranta, and Salo 2022). Additionally, unlike traditional methodologies, which are susceptible to errors, Python's capacity for automation and scalability enables the efficient processing and analysing of large volumes of textual data. Furthermore, Python's status as an open-source platform and there is active support from a dynamic community, offer researchers an avenue to access the latest techniques, hereby bolstering data analysis capabilities (Bochkay et al. 2023).

We use the wordlist developed by Mansouri and Momtaz (2022), which specifically identified distinct words associated with each of the three components of ESG. After cleaning the text of the corporate narrative and employing requisite Python libraries, we compute the frequency of occurrences for each word list in the corporate narrative. In particular, we analyse each university's annual report by searching and counting the specific words that relate to ESG. These word counts are guided by topic-specific dictionaries, as outlined in the methodology developed by Mansouri and Momtaz (2022).

Our independent variable is the VCs' career horizon. Following Strike et al. (2015), a VC's career horizon is defined as 70 deducting a VC's current age. We select 70 years of age as the end of the VC's career based on the assumption that the VC's retirement age is around 70 years. In other words, an older VC has a relatively shorter career horizon, whereas a younger VC has a relatively longer career horizon.

To explore the moderating effect of soft information, we introduce the following two variables, including ‘boilerplate score’ and ‘forward-looking’ information. Specifically, to obtain the boilerplate score of each observation, we initially define boilerplate phrases as those 4-g that are more commonly used than average in the reports of that year. Subsequently, we calculate the boilerplate score as the total number of boilerplate phrases in annual reports exceeding the average frequency observed within a given year, divided by the total number of 4-g present in those annual reports (Küster 2024). In addition, we define ‘forward-looking’ information as the ratio of sentences that reference future activities or expectations to the total number of sentences in annual reports based on Loughran and McDonald (2015). This metric quantifies the extent of forward-looking information provided in the university annual report disclosures.

A series of control variables are also included in our model, including a VC's other characteristics, university-specific information, and the financial performance of the university. The information on the VCs' other characteristics includes their education level and compensation. In particular, VC education is defined as a dummy variable that equals one if a VC has an environmental or sustainability related educational background or degree, and zero otherwise. It is generally believed that a VC with a sustainability-related degree is more likely to pay attention to a university's ESG compliance. In addition, we control a VC's compensation, including their base salary, pension, and other short- and long-term benefits. This is because a VC may over-invest in ESG-related activities and push for more sustainability disclosures to build their personal reputations and secure greater compensation (Jian and Lee 2015).

In terms of university-specific information, we control ‘student-to-staff ratio’, ‘Big 4’ and ‘Post 1992 University’ status. Firstly, we include the ‘student-to-staff ratio’, recognising that a lower ratio indicates more available staff resources. This increased staff resource capacity within the university is expected to help enhance sustainable performance. Additionally, we incorporate a control variable for ‘Big 4’ accounting firms, which is a dummy variable that equals to one if a university's financial statement is audited by one of the biggest four accounting firms, and zero otherwise. Previous evidence such as Pflugrath, Roebuck, and Simnett (2011) find that ‘Big 4’ accounting firms are more likely to provide reliable ESG assurance services. Moreover, we introduce a control variable named ‘Post-1992 Uni’ represented as a dummy variable. It takes the value of one if a university is a member of the Post-1992 University Group, and zero otherwise. Generally, post-1992 universities are more entrepreneurial-focused and have relatively restricted financial resources compared to Russell Group Universities (Fotiadou 2022). Consequently, this may limit their ability to invest in ESG-related activities.

Regarding university financial performance, we include the following financial variables such as university size, leverage and return on assets. We include university size as universities with greater size are in a better position to allocate resources, invest in and actively participate in sustainable activities (Huang et al. 2023). Universities with higher leverage are more prone to prioritising short-term financial interests, and therefore are less likely to pursue sustainability-related disclosures (Wang et al. 2024). Subsequently, we include leverage as a control variable. Lastly, we include return on assets, which is calculated as the ratio of net income to total assets. This is attributed to the likelihood that universities with stronger financial performance are more inclined to allocate resources and carry out sustainable related initiatives (Huang et al. 2023). The variable definitions are summarised in Table 1.

3.3 Research Method

To examine moderating Hypothesis 2a, we reassess Equation (1) by segregating the entire sample into two subsamples: one comprising university annual reports with higher boilerplate scores (i.e., scores higher than the sample median), and the other with lower scores. Similarly, to examine Hypothesis 2b, we divide the entire sample into two subsamples: university annual reports with greater forward-looking information (i.e., a ratio greater than the sample median), and those with less forward-looking information.

4.1 Descriptive Statistics

The summary statistics are presented in Table 2A. On average, the sustainability score of the sample universities is 484.28. Additionally, the mean VC's career horizon is about 10.59 years. About 6% of observations have boilerplate phrases exceeding the average frequency observed within a given year. Moreover, approximately 1% of words refer to future activities in the universities' annual reports. Moreover, approximately 13% of VCs have sustainability and environmental-related degrees, and on average, they earned about £298,107. It is also reported that approximately 64% of university annual reports are audited by ‘Big 4’ accounting firms. The correlation matrix has also been calculated. According to the results in Table 2B, it indicates that multicollinearity is not a concern for our study.

4.2 Main Results

Table 3 presents results for H1. In Model 1, the dependent variable is each university's ESG score, and in Models 2–4, we decompose ESG scores into Environment (E), Social (S) and Governance (G) scores respectively as alternative dependent variables and run them against the independent variable of VCs' career horizon. It is observed that the coefficients of VC career horizon are significantly negative across Models 1–4. Given our definition of career horizon (e.g., 70—VC current age), this finding can translate to the fact that older VCs tend to pay greater attention to the university's sustainability performance. This is probably because a younger VC may tend to focus more on financial performance and thus pay less attention to sustainable performance at an early stage. When they get older, financial concerns reduce, and they are more likely to focus on addressing sustainability issues. It resonates with previous findings indicating that older professionals tend to exhibit greater ethical and conservative behaviour, addressing social and environmental concerns (Peterson, Rhoads, and Vaught 2001).

Models 1 and 2 of Table 4 present the impact of VC career horizon on university sustainability performance by higher and lower boilerplate scores separately. It is observed that the relationship between career horizon and university sustainability performance is significantly negative in the high boilerplate score group. In comparison, there is no impact of career horizon on sustainability performance in the low boilerplate score group. This finding supports Hypothesis 2a, indicating that longer VCs' career horizon deteriorates university sustainability performance, especially when universities strategically use more repetitive, vague and uninformative content to fill their annual reports.

Models 3 and 4 of Table 4 report results for Hypothesis 2b. For universities that use more forward-looking statements in their annual reports, the estimated coefficient of career horizon in Column (3) is significantly negative. In contrast, there is no relationship between career horizon and sustainability performance when universities use fewer forward-looking statements in their annual reports. These findings suggest that when universities have high sustainability concerns, they use more forward-looking statements to dilute the focus on current performance. Subsequently, the greater number of forward-looking statements in a university's annual report, the more pronounced the negative impact of VCs' career horizon on university sustainability performance.

Our findings are embedded within the framework of UET. According to UET, top executives' observable characteristics significantly influence organisational outcomes (Hambrick and Mason 1984). Our findings align with this theory by demonstrating that VCs, akin to CEOs, exert substantial influence on the sustainability initiatives of HEIs, particularly through their career stage. This is consistent with previous research (e.g., Wang et al. 2024) that has utilised UET to examine the impact of CEOs on environmental activities. Specifically, for VCs nearing retirement, they are more likely to drive substantive sustainability initiatives, reflecting a deeper ethical commitment and strategic focus on long-term environmental goals. In this regard, we contribute to the growing body of literature that highlights the role of VC characteristics in shaping institutional performance (e.g., Elmagrhi et al. 2021).

Moreover, our results highlight the importance of transparency in sustainability reporting, as advocated by Cheah et al. (2023), who examined VC power and HEI efficiency. We observe that the use of non-substantive, boilerplate and forward-looking language in sustainability reports undermines the credibility of sustainability efforts, suggesting a need for more genuine and transparent communication. This finding contributes to the literature by emphasising the important role of clear and honest ESG reporting in enhancing the effectiveness of sustainability initiatives.

4.3 Robustness Tests

We have conducted several additional analyses for robustness purposes. First, we consider if our results are robust to alternative choices of dependent variables. Subsequently, we replace ‘sustainability performance’ with ‘sustainability score’ and ‘SDG 17’. These two variables are obtained from the Times Higher Education Impact Ranking official website. The ‘sustainability score’ examines how a university manages its sustainable performance. In addition, the ‘SDG 17’ emphasises the implementation and revitalisation of the Global Partnership for Sustainable Development. This is measured as a dummy variable that equals one if a university's annual report includes information relating to SDG 17, and zero otherwise. The results reported in Table 5 (Models 1 and 2) display statistical significance, indicating that the VCs' career horizon reduces sustainable performance, which is consistent with our previous findings.

We then re-estimate the main regression analysis using another regression model: Poisson regression. This is because our dependent variable is a count of the number of disclosures. Poisson regression is particularly suitable when the outcome variable represents the number of occurrences of an event within a fixed period (Coxe, West, and Aiken 2009). The results are reported in Table 5 (Models 3–6), indicating that our main findings remain unchanged even when selecting alternative regression models.

We also examine how a VC's gender affects the relationship between their career horizon and sustainability performance. Generally, females value social responsibilities highly relative to their male counterparts (Elamer and Boulhaga 2024; Elamer, Boulhaga, and Ibrahim 2024; Hui, Li, and Elamer 2024). Consequently, female VCs may exert additional pressure on universities to engage in socially responsible activities (Wang et al. 2022). Models 1 and 2 of Table 6 report the impact of the career horizon on sustainability performance among female and male VCs, respectively. It is reported that there are significant and negative coefficients in both groups, although the magnitude of the coefficient is greater in the female VC group. The results indicate that older female VCs with shorter career horizons pay more attention to sustainability activities, leading to enhanced sustainability disclosure by universities.

Next, we examine the relationship between career horizon and sustainability performance while considering the university's ‘Russell Group Status’. The Russell Group is a self-selected association of 24 public research universities in the United Kingdom, known for their academic excellence and research-intensive focus (Elmagrhi et al. 2021). These universities are considered more prestigious and are therefore more likely to receive greater financial support. In this regard, we may expect that in these universities, longer career horizon VCs may not necessarily reduce sustainability performance, as younger VCs with longer career horizons have fewer short-term financial concerns. The empirical results presented in Models 3 and 4 of Table 6 indicate that there is a negative relationship between career horizon and sustainability performance in the non-Russell Group universities. In contrast, there is no impact of career horizon on sustainability performance for universities belong to the Russell Group, which is in line with our aforementioned expectation. We also control for the four countries (England, Scotland, Wales and Northern Ireland) in Model 5 of Table 6, and the results remain consistent, confirming the robustness of our findings.

To address potential endogeneity arising from differences in fundamental characteristics between the treated and control groups, we employ propensity score matching (PSM). This non-parametric method allows for a ‘like-for-like’ comparison, crucial for examining the relationship between VCs' career horizon and university sustainability performance (Noureldeen et al. 2024; Owusu et al. 2023; Ullah, Owusu, and Elamer 2024). Our analysis involves matching 154 universities in the treatment group with 350 in the control group, using data derived from university annual reports that include adherence to the UN (SDGs).

The results, detailed in Models 1–5 of Table 7, affirm the consistency of our findings with our primary results displayed in Table 3. We observe that universities with VCs nearing the end of their career horizon tend to exhibit stronger sustainability performance. This pattern underscores the influence of VCs' career horizon on sustainability strategies, with older VCs demonstrating a stronger commitment to sustainability objectives. Such findings highlight the critical role of leadership characteristics in shaping university strategies towards sustainability, reinforcing the robustness of our analysis against potential biases introduced by unobserved heterogeneity. Overall, our PSM approach not only addresses the self-selection bias effectively but also supports the robustness of our findings.

Lastly, we explore the mechanisms through which VCs' career horizons impact university sustainability performance in Table 8, where channel analysis provides insights into the dynamic interplay between governance structures and leadership tenure. This table highlights the significant role of council meetings and sustainability committees in moderating the influence of VCs' career horizons on sustainability outcomes. Firstly, the interaction between VCs' career horizons and the frequency of council meetings suggests that more engaged governance, through frequent council meetings, can mitigate some of the adverse effects associated with longer career horizons. Specifically, Model 1 of Table 8 shows that the coefficient on the interaction term between horizon and council meetings is negative and significant, indicating that as VCs near the end of their careers, the presence of active governance frameworks can still foster a commitment to sustainability. This finding underscores the importance of robust governance structures that can support sustained organisational focus on sustainability goals, irrespective of individual leadership timelines.

Secondly, the analysis involving sustainability committees reveals a more pronounced effect. The presence of these committees significantly enhances the sustainability performance of universities, as shown in Model 2 of Table 8 by the positive coefficient for sustainability committees alone. However, the negative interaction term between career horizon and sustainability committees suggests that while sustainability committees generally promote better sustainability practices, their effectiveness is somewhat reduced when VCs are closer to retirement. This could be interpreted as a need for these committees to have stronger empowerment or more independent functioning, especially when dealing with VCs who might have shorter-term focuses towards the end of their careers.

Overall, Table 8 highlights that while the career horizon of VCs does influence sustainability performance, this effect can be moderated by institutional mechanisms like council meetings and sustainability committees. These results suggest pathways for enhancing university sustainability practices through strategic governance adjustments, particularly in light of leadership transitions. These insights are crucial for policy-makers and university boards aiming to enhance their sustainability trajectories in times of leadership change.