Introduction

Research-intensive universities around the world have multiple goals, being tasked with contributing to multiple missions and complementary objectives (Miller et al., 2021; Morphew et al., 2018). They are responsible for teaching, generating knowledge through research, promoting scientific development and addressing social and economic issues (Altbach, 2011; Eidt and Calgaro, 2021; Li et al., 2024). Universities play a crucial role in societal development and serve as democratic institutions of general interest to a country or region (Balbachevsky, 2015; Petersen, 2025). A balanced combination of teaching, research, and outreach missions has been considered essential for achieving excellence in academic quality (Nicotra et al., 2021).

Theoretical perspectives on university evaluation, particularly in resource allocation, have advanced significantly in recent decades. Thomas et al. (2020) emphasises that the integration of objective criteria and systematic processes is essential for promoting academic excellence and quality. Furthermore, the approach of performance-based resource allocation, which incentivizes academic units to enhance quality, provides foundations for understanding the link between budget distribution and institutional outcomes.

The key characteristics of research universities in the 21st century, including a global mission, intensive research, diversified funding sources, worldwide recruitment, increasing complexity, new relationships with government and industry, international collaborations and national socioeconomic development (Mohrman et al., 2008; Mochnacs et al., 2024). Morphew et al. (2018) and Moscardini et al. (2020) noted that public research universities, given their multiple functions, are more likely to be exposed and required to demonstrate their results as they use public resources. Therefore, effective management of research universities is increasingly important in an era of complex and eminent academic organisations (Trakman, 2008).

According to Zechlin (2010) and Gede and Huluka (2023), effective implementation of planning can significantly contribute to the progress and transformation of universities. Planning can help influence an organisation’s strategic direction through coordination and effective decision-making. Scholars have provided evidence of the need and feasibility of strategic planning in universities, demonstrating positive outcomes for those who adhere to planning (Fumasoli and Lepori, 2011; Nguyen and Van Gramberg, 2018; Falqueto et al., 2020).

Over time, universities have transitioned from being a community of scholars to an organisation with multiple stakeholders (Bleiklie and Kogan, 2007; Trakman, 2008; Douglass, 2016; Egorov and Serebrennikov, 2023; Petersen, 2025). Universities’ success now largely depends on their organisational and managerial skills (Meek et al., 2010). Although high-performing professors, academics, and students are pivotal for their contributions and reputation, modernising and professionalising their management models and capabilities have become increasingly important (Bleiklie and Kogan, 2007; Jones et al., 2021).

Similarly, Nguyen and Van Gramberg (2018) and Egorov and Serebrennikov (2023) emphasised that universities require capable management to organise and structure their activities effectively and have qualified teaching and research professionals. In a context where administrative structures must manage highly autonomous constituent parts (Douglass, 2016), budget allocation can serve as an incentive for achieving the institution’s strategic objectives. Hinton (2012) highlighted the importance of long-term budgets aligned with the university’s strategic planning and goals and explicitly referred to the institution’s objectives. Powell (2017) suggested that resource allocation and institutional strategic objectives must be linked to help universities become more competitive.

However, the existing literature lacks sufficient focus on strategic planning and budgeting in research universities, particularly in the context of resource allocation. While some studies have recognised the importance of improving resource allocation efficiency, there remains a gap in understanding how these processes are implemented within specific types of universities, especially in the Global South. Chen et al. (2024) highlight the gap in research on the resource allocation efficiency of top-tier universities in China, noting that while there is a general understanding of the need for efficient resource management, the existing studies fail to provide a clear, purposeful analysis of these universities. This gap underscores the need for more in-depth investigations into how strategic planning and budget allocation are linked to institutional objectives such as academic excellence and socioeconomic relevance. To address this gap, this article adopts a case study methodology to explore these themes in a specific context, aiming to contribute to the literature by providing insights into budget allocation models that align with academic and strategic goals in the Global South (Cantwell et al., 2022).

Although existing literature underscores the significance of strategic planning and resource allocation in research universities, critical gaps persist. Studies emphasise the need for efficient resource management (Zechlin, 2010; Hinton, 2012), particularly in the Global South, but often lack a detailed examination of how these processes operate within specific institutional contexts. A significant portion of research focuses on broader regional or national frameworks (Chen et al., 2024; Hoareau et al., 2013) rather than the intricacies of resource allocation within individual universities. Moreover, while the correlation between budget allocation and academic performance is well-established (Morgan Jones et al., 2022), there is a dearth of analysis on how resource allocation models directly shape strategic objectives such as academic excellence, innovation, and socioeconomic impact.

The newly proposed model presents a multidimensional and adaptable framework for resource allocation in research universities. It integrates strategic planning, social impact, sustainability, and inclusion, ensuring that budget allocation aligns with long-term institutional goals. A key innovation is the dynamic weighting of indicators through the Analytic Network Process (ANP) and PROMETHEE methods, allowing flexibility across different academic disciplines. The model also enables institutional customisation, permitting universities to prioritise indicators according to their specific missions and regional contexts.

By incorporating mathematical decision-support tools, the model ensures that funding decisions are data-driven and context-sensitive, overcoming the limitations of conventional allocation mechanisms. This research offers both a theoretical and practical contribution, providing universities with a decision-making tool to enhance institutional excellence and strategic coherence.

This study contributes to bridge these gaps by investigating strategic planning and resource allocation at a prominent Latin American research university. The methodology offers insights into how universities align their budgeting practices with institutional priorities.

Therefore, the subsequent sections present the following: (i) a literature review on planning and budget allocation in research universities, which also includes case studies from universities in several countries; (ii) the methodology and data used to analyse the case of a leading research university in Latin America and its main results; (iii) a newly proposed resource allocation model that integrates multicriteria decision-making methods to achieve strategic objectives in the university’s continuous pursuit of excellence.

Prior literature: budget planning, strategy, and management

To review the literature on strategic planning and resource allocation in research universities, we selected the Web of Science and Scopus databases due to their comprehensive coverage of the relevant topics (AlRyalat et al., 2019; Adriaanse and Rensleigh, 2013). The Bibliometrix software was used as a web interface to facilitate data import and filtering (Aria and Cuccurullo, 2017).

To select the sample, the parameter involved searching for publications that matched the commonality of three-term groups: strategic planning (group 1), resource allocation (group 2), and universities (group 3). The terms used are listed in Table 1.

Table 1 Search terms used in the Web of Science and Scopus databases.
Full size table

After removing duplicates, the data were exported to Bibliometrix to identify thematic clusters. A minimum of five joint occurrences between terms was maintained to create these clusters, and the first 290 keywords were considered. Subsequently, a diagram was applied to display the themes on the centrality and density axes (Aria and Cuccurullo, 2017), as shown in Fig. 1. Clusters that intersect the axes exhibit good centrality or publication density, while those in the first quadrant demonstrate good density and centrality. However, the clusters ‘resource allocation’, ‘allocation strategy’, and ‘decision-making’ are located in the third quadrant, indicating low density and centrality of publications on these topics.

Fig. 1
figure 1

Clusters articles in centrality and density axes plotted, using Bibliometrix.

Full size image

Some of the studies found in the search are highlighted below, whose approaches include strategic planning and resource allocation in universities of interest to this article.

Some studies analysed the correlation between budget management and the pursuit of strategic objectives, which encompass expanded indicators of quality and excellence. A well-known example of institutional financial incentives and academic performance (broadly encompassing the three objectives mentioned above) is the Research Excellence Framework (REF) in the UK. Various studies have analysed the impact of the REF on universities (Morgan Jones et al., 2022; Oancea, 2019).

Van Noorden’s (2015) analysis of 7,000 case studies from the UK’s REF reveals that projects containing terms such as ‘million’, ‘market’, ‘government’, ‘principal’, and ‘global’ are highly correlated with better evaluations. Oancea (2019) emphasised that evaluation has become central to the political dynamics of relationships between universities, society, and the state, with academic quality now being analysed alongside the broader impacts of research developed by universities.

Hoareau et al. (2013) conducted a comparative analysis of higher education policies in 32 European countries, and found that an increase in university autonomy and public funding positively impacts universities’ research and educational performance. It also enhances countries’ innovation potential, leading to a positive effect on their economies (Li et al., 2024).

Bisogno et al. (2014), who examined the financial situation of Italian universities, concluded that financing policies should incorporate methods to monitor their indicators and evaluation criteria, focusing on costs and the quality of services provided. Watermeyer and Chubb (2019) stated that by including the research’s economic and social impact as a performance indicator and evaluation, the criterion had been a game-changer for the REF in the UK, transforming how universities act. This also challenged perceptions and practices of research excellence, resulting in a new form of academic distinction based on economic and social impact (Watermeyer and Chubb, 2019).

The theoretical framework is constructed based on Douglass’ (2016) definition of new flagship universities, which are research-intensive institutions with broad goals, being challenged to define and redefine their objectives and significantly expand their societal role. This model does not neglect international standards of excellence, which are largely focused on research productivity, but emphasises national and regional services with specific characteristics and responsibilities beyond those considered in traditional international rankings (Douglass, 2016; Chun and Sauder, 2023). Management methodologies oriented towards augmenting this diversity may be deployed as instruments for augmenting university efficiency and performance (Egorov and Serebrennikov, 2023).

Hinton (2012) highlighted that information for various budgetary aspects does not originate from a single source, and strategic planning is necessary to integrate it. The authors argue that this comprehensive context is essential to ensure the appropriate allocation of budgetary resources to support the institutional objective (Hinton, 2012).

The success of a public institution can also hinge on how adequately it plans and executes its budget (Falqueto et al., 2020). Therefore, implementing a strategic budget management system can enable universities to go beyond staff remuneration and maintenance expenses. Instead, they can drive new initiatives and undertake strategic projects that enhance the quality of education and fulfil their objectives. Although not an absolute requirement, it is one of several management actions that can contribute to achieving this goal.

These studies demonstrate budget management’s significant role in developing university goals. However, they do not provide a detailed analysis of how budget allocation operates as either incentives or disincentives for teaching, research, and extension. To address this gap, this study examines the allocation model adopted by University of Campinas and proposes an original alternative model that aims at contributing to the topic.

In addition, several recent studies have provided further insights into the evolving challenges faced by universities, particularly in the context of learning and sustainability (Zhu et al., 2024). Elbawab (2024) indicates that universities are undergoing changes that can be addressed through learning, highlighting the positive relationship between organisational learning culture and organisational performance. This is further supported by the evidence suggesting that organisational learning helps universities achieve better performance and sustainability. Additionally, organisational learning is shown to reduce uncertainty, as pointed out by Elbawab (2024), emphasising its importance for enhancing university efficiency and adaptability.

For the development of sustainability in higher education, the works of Zhu et al. (2024) and Elbawab (2024) emphasise that it can be implemented in teaching, research, governance and extension, aligning with the strategic objectives of universities.

Considering factors that enhance research productivity and innovation, Javed et al. (2024) examine the resource allocation in South Asian universities. They find that this topic significantly influences academic research efforts and is essential for promoting academic progress and innovation (Javed et al., 2024).

Still considering resource allocation efficiency, Chen et al. (2024) investigated 13 top-tier universities in China. They could confirm that factors such as the regional economic environment, faculty structure, and international exchange play key roles in promoting resource allocation efficiency. Additionally, they emphasise inherent complexity of these processes and their impact on institutional performance (Chen et al., 2024). These studies collectively underscore the critical role of resource allocation, as explored and proposed in this work, including strategic planning in improving university purposes.

Case of University of Campinas

The University of Campinas consistently ranks as one of the top three universities in Latin America. The study of this descriptive-analytical case aims to explore evaluation practices. According to Thomas et al. (2020) definitions, the object of investigation comprises the administrative means to promote excellence and quality. Meanwhile, the university engaged in intensive research activities is the subject of study. Established in 1966, it is a public university with 24 academic units (listed in Appendix 1) and circa 35,000 enrolled students across 66 undergraduate courses and 153 graduate programmes. Its annual budget, which comes from taxpayers in the state of São Paulo, reached approximately US$700 million in 2021. Within the constraints set forth by Brazilian regulations, the university enjoys a degree of autonomy in determining how to allocate this budget.

Since 1995, the budget allocation has been partially determined by the Budget Qualification Programme (BQP), which comprises two subprogrammes: Undergraduate Teaching Support Programme (Teaching-BQP) and Research Productivity Quality Support Programme (Research-BQP). The central administration implemented BQP based on the assumption that a variable and rewarding budget could positively influence academic performance. Table 2 displays the criteria used in both subprogrammes.

Table 2 Criteria used in the two subprogrammes of the current BQP (Budget Qualification Programme).
Full size table

Upon analysing the current criteria presented in Table 2, it is evident that the subprogramme focused on teaching prioritises the number of students and workload in undergraduate studies, with a double weight given to the number of students in the evening period. The research-focused subprogramme prioritises the number of faculties with a PhD degree, master’s and doctoral candidates, and publications. These traditional indicators essentially measure the weight of teaching, especially in the night period, and the production of graduate students and publications. The criterion of the number of professors with a PhD degree should have been excluded several years ago because the entire faculty members of University of Campinas are composed of PhD holders. Furthermore, maintaining these criteria contradicts recent discussions on evaluation and impact (Thomas et al., 2020) as they do not include qualitative indicators or indicators focused on extension, culture, and innovation.

Assessing the efficacy of BQP in promoting academic performance

The descriptive analysis of the budget distribution during the mentioned period revealed that, on average, 46% of the ordinary budget of the 24 academic units (listed in Appendix 1) was sourced from BQPFootnote 1. The evaluation was divided into three time intervals to assess the impact of budget allocations over time (1995–2003; 2004–2014; 2015–2021).

Figure 2a illustrates the evolution of budget allocation resulting from BQP criteria for each academic unit on the x-axis and the corresponding fraction received from BQP during the first analysed period (1995–2003) on the y-axis. Colours have been employed to distinguish between units in the three analysed periods. Some units did not exist during this period and are at zero on the y-axis. In Fig. 2b, the box plot of allocated values for the second analysed period (2004–2013) showed a reduction in variation amplitude compared to the previous period. In the third analysed period (2014–2020), as depicted in Fig. 2c, despite the entry of new units, which generated variation in the allocated fraction of others, the variation amplitude remained significantly lower than in the first two analysed periods.

Fig. 2: Box plot of the BQP fraction per academic unit for three distinct periods.
figure 2

a 1995–2003, b 2004–2014, and c 2014–2020.

Full size image

The analysis of the three periods revealed that the programme’s amplitude varied, on average, by 0.022 units during the first period, decreased to 0.014 in the second period, and stabilised in the last period with a value of 0.010. This indicates that the programme has achieved stability over time.

An analysis of the correlation between the fraction received by the units for each subprogramme has revealed a strong correlation with only one of the considered indicators. In the Teaching-BQP, the teaching workload has been the most significant criterion over the last seven years, with an average of 0.92. Similarly, in the Research-BQP, the number of teachers in the units has been the most important criterion, with an average of 0.87. The other criteria have had a marginal influence on the distribution of resources, with a correlation below 0.50. This result indicates a loss of BQP’s ability to incentivise quality.

Furthermore, the efficiency frontier of BQP was analysed through DEA (Gregoriou and Zhu, 2006) and SFA (Aigner et al., 1977), and both methodologies yielded consistent results, with many units located on or near the efficiency frontier. This suggests that there are no inefficiencies in resource allocation and that, in the long term, BQP has ceased to promote indicators of quality in academic units.

Methodology

Data on budget allocation across the twenty-four academic units from 1995 to 2020 were collected from statistical yearbooks and budget proposals (AEPLAN, 2022). The analyzes include box plot analysis for three different periods and correlation analysis of current indicators with the fractions allocated in the two subprogrammes of BQP. Subsequently, the impact of BQP on the performance of academic units was evaluated using the indicators employed by the programme itself. Two techniques, Data Envelopment Analysis - DEA and Stochastic Frontier Analysis—SFA, were used for this purpose. The DEA method (Gregoriou and Zhu, 2006) assesses the relative efficiencies of units handling multiple inputs and outputs, employing non-parametric data in frontier analysis. SFA (Aigner et al., 1977) evaluates the relative efficiency and performance of units by quantifying their proximity to optimal efficiency. The main objective was determining the extent to which BQP motivated units to improve their indicators to receive a larger share of the central budget.

After testing the effectiveness of BQP in enhancing academic performance, as measured by the programme’s indicators, among the academic units of University of Campinas, it was found that this model turned ineffective overtime, as discussed in the next item. Therefore, a new model based on two main premises is proposed here. The first premise emphasises the need for more comprehensive indicators that align with the objectives of research universities, as discussed earlier. The second premise focuses on calculating this set of indicators, which should be sufficiently flexible to accommodate the diversity of situations and indicators of the different areas of knowledge present in research universities.

Building on the theoretical frameworks of Behzadian et al. (2010) and Saaty (2004), this study employs a methodology that captures both the complexity and diversity of performance metrics in research universities. The PROMETHEE method (Behzadian et al., 2010) enables nuanced evaluation by integrating qualitative and quantitative factors, while the ANP methodology (Saaty, 2004) accounts for interdependencies across multiple dimensions, reflecting the interconnected nature of academic performance.

The construction of the academic performance evaluation model related to budgetary incentives began with a broad range of indicators obtained through a literature review and aligned with the strategic objectives of University of Campinas’s planning. The model adopted a performance analysis method based on multicriteria outranking methodology due to its flexibility in defining weights and preferences. The PROMETHEE method—Preference Ranking Organisation Method for Enrichment Evaluations (Behzadian et al., 2010) was used, which can be applied to various alternative classification problems in the presence of multiple criteria. The proposal of this work to use PROMETHEE in association with ANP—Analytic Network Process (Saaty, 2004) for resource allocation between teaching and research units includes the following advantages: objectivity in the necessary data; easy obtainment; reduced computational complexity; agility in interpretation; identification of strengths and weaknesses. Further, it is not a deterministic method, allowing the use of stochastic data. Moreover, PROMETHEE allowed for the verification of the net flow result (Phi), as well as the separate result of the strengths and weaknesses points in the performance of each unit (positive flow, Phi+; and negative flow, Phi-) in relation to the other units (Goumas and Lygerou, 2000; Pohekar and Ramachandran, 2004).

Several simulations were conducted using the currently available indicators. Initially, a PROMETHEE simulation was performed with equal weights assigned to the institutes and faculties. Subsequently, simulations were conducted by varying the weights of the 18 indicators, with the units grouped by area of knowledge, and displaying the positive flow result to visualise the outcome based on each unit’s strengths. Finally, the results were simulated for each unit to visualise the performance of each indicator.

The proposed methodology augments budgetary incentives for excellence through the incorporation of expanded criteria and adaptable calculation techniques that accommodate diverse knowledge domains. This approach comports with global research university standards, thereby ensuring a modernised and pertinent allocation framework.

Weight assignment within the Analytic Network Process (ANP) method was executed by constructing a network of interdependencies among the proposed dimensions and their constituent indicators. These dimensions were hierarchically structured based on the university’s strategic objectives and expert appraisal. ANP facilitates the comparison of the relative significance of dimensions and indicators through pairwise assessments, thus determining weights accordingly. Consistency ratios were computed to validate judgments, with requisite adjustments made to ensure a precise and context-sensitive distribution of weights across academic units.

The selection of ten dimensions was predicated upon University of Campinas’s strategic objectives and pertinent scholarly literature (Morgan Jones et al., 2022; Bandola-Gill et al., 2021; Moed, 2020). Each dimension encapsulates a key facet of university performance, enabling a comprehensive evaluation of both academic and non-academic outcomes. The dimensions were also chosen to reflect the multifarious roles of universities, including societal engagement, environmental sustainability, and inclusion, which are increasingly pivotal to institutional success. The decision to utilise ten dimensions constituted a simplified categorisation paradigm rather than a rigid stipulation.

Indicator thresholds were established based on historical data, expert consultation, and extant standards in higher education performance evaluation. These thresholds afford flexibility, adapting to varying institutional contexts and priorities. They can be recalibrated over time in response to performance trajectories and evolving university administrative imperatives.

The proposed methodology aims to modernise the budgetary stimulus towards excellence by incorporating broader and more diverse criteria and flexible calculation techniques that can account for the diversity of knowledge areas. This aligns with the demands placed on research universities worldwide and ensures that the stimulus is based on current standards.

Results

Proposal for a new model

The proposed model distinguishes itself by incorporating a broader and more dynamic range of dimensions. Beyond conventional indicators, it integrates aspects such as engagement with society, inclusion and diversity, and environmental sustainability, with particular emphasis on the local and regional impact of universities, recognising their expanding role in addressing global challenges.

Additionally, the application of the Analytic Network Process (ANP) method for assigning variable weights to indicators based on diverse fields of knowledge enables a more precise and tailored approach. This methodology enhances flexibility and stability in evaluation, effectively capturing the complexities of contemporary universities and their multifaceted responsibilities. Consequently, the model offers a significant contribution to the literature by providing a comprehensive and adaptable framework that more accurately reflects universities’ strategic objectives and their broader social impact.

After conducting various analyses and discussions, this article introduces a resource allocation model using the University of Campinas as a case study. The model incorporates the strategic objectives outlined in the university’s planning process, which theoretically ensure that the proposed indicators align with the institution’s medium and long-term priorities. Table 3 presents these objectives.

Table 3 Strategic objectives of University of Campinas Planning.
Full size table

The university’s strategic objectives encompass diverse topics relevant to the three goals of research universities. These topics are reflected in the literature cited in the introduction and literature review, particularly in the studies of Morgan Jones et al. (2022), Bandola-Gill et al. (2021), Moed (2020), Thomas et al. (2020), Oancea (2019), Joly and Matt (2017) and Falqueto et al. (2020).

These studies underscore the importance of flexible performance metrics, particularly emphasising the central role of evaluation in universities for their relationship with society and the government (Oancea, 2019). They also highlight the following: the significance of the broader impacts of research developed by universities (Morgan Jones et al., 2022); the need for institutions to conceptualise and implement internal quality control and funding procedures (Moed, 2020); the growing focus on socioeconomic impact as a criterion for research funding evaluation (Bandola-Gill et al., 2021); the search for effectiveness and efficiency beyond self-reports and local impacts (Thomas et al., 2020); the importance of diversity in impact assessment tools (Joly and Matt, 2017).

Based on the strategic objectives of the university and a review of the current literature on the topic, Appendix 2 presents ten dimensions and their respective indicators as the foundation for the proposed model’s criteria in this study.

The quality of education dimension is measured by various indicators, such as the number of undergraduate and graduate students, implementation of new learning techniques, scholarship for students, and the workload completed in laboratories and extension activities.

The research dimension encompasses the Field-Weighted Citation Impact (FWCI) indicator or similar indicators that consider differences in fields of knowledge. The FWCI measures the number of citations received by a university’s publications compared to the average number of citations from similar publications. Moreover, the research dimension includes the following: Altmetrics, research projects conducted in collaboration with other research institutions, international research collaborations, Number of researchers linked to the National Research Council, publications in partnership with companies and startups, intellectual property registrations, obtained licences, number of citations of articles in patents and other intellectual property instruments, and citations of articles in public policies.

The engagement with society dimension encompasses the number of students enrolled in extension courses, community activities, and the organisation of scientific, cultural, and artistic events.

The financial sustainability dimension encompasses the funds gathered through extension activities, technical and institutional reserve resources of graduate studies, resources acquired through contracts and agreements, and resources received from funding agencies.

The following indicators measure internationalisation: the number of exchange students per enrolled student during the daytime period; the proportion of foreign undergraduate students to total enrolled students; the proportion of foreign graduate students to total enrolled students; the participation of foreign co-authors in publications.

The inclusion and diversity dimension encompasses indicators of socioeconomic equity (undergraduate students from public schools), racial equity (black and brown people), and gender equity (women compared to men in graduate students).

Visibility, as a dimension focused on communication (Petersen, 2025), includes website and social media visits, press mentions both within and outside the university, and visits and downloads on the university library system’s thesis and dissertation repository.

The administrative efficiency dimension is dependent on the computerisation, mapping, and optimisation of processes; the environmental sustainability dimension encompasses energy and water consumption, as well as carbon balance. Finally, the quality of life at work dimension encompasses various indicators, such as communal, reception, and referral spaces, and the promotion of internal cultural activities.

After obtaining the dimensions and their corresponding indicators through the strategic objectives of the university and relevant literature, the subsequent step was to propose a method for assigning weights to these indicators, as they can (and should) have varying weights across different areas of knowledge (Thomas et al., 2020; Moed, 2020).

We used the ANP—Analytic Network Process method (Saaty, 2004) to assign weights. This technique is particularly useful for non-hierarchical problems, as it allows for interactions and dependencies between elements at distinct levels, resulting in greater stability and more precise modelling of complex situations (Mikhailov and Singh, 2003; Jharkharia and Shankar, 2007). ANP also provides reliability, allowing for measuring the internal consistency of stakeholder judgements, which is crucial when considering the diversity of knowledge areas in research universities. Additionally, ANP can be associated with other multicriteria methods, such as the PROMETHEE (Preference Ranking Organisation Method for Enrichment Evaluations) outranking method (Behzadian et al., 2010), which will be incorporated into the model.

Figure 3 illustrates the ANP structure with the proposed dimensions and indicators in this study. The Super Decisions software generated the image (Liu et al., 2003). The ten dimensions (lines two and three of Fig. 3) are interconnected and aligned with the university’s strategic objectives (line 1 of Fig. 3). The attainment of these objectives relies on the outcomes achieved by the university’s institutes and faculties (line 4 of Fig. 3).

Fig. 3: ANP —Analytic Network Process of the proposed dimensions.
figure 3

Novel proposed model: structure of the ANP with the dimensions. The ten dimensions and their indicators interact with each other in search of the university’s strategic goals that go through the results of the institutes and faculties.

Full size image

After assigning weights through ANP, the data are entered into PROMETHEE, which generates a ranking of academic units by area of knowledge. PROMETHEE also allows for visualising each faculty and institute for the selected indicators. In a simulation, the results were obtained in two main ways: (i) by academic units per area of knowledge (Figs. 4 and 5a–e); (ii) by the performance academic units using an example of two from the interdisciplinary area (Fig. 6a, b).

Fig. 4
figure 4

Overall outranking of units in PROMETHEE, equal weights for indicators, result for Phi = Phi(+) + Phi(-).

Full size image
Fig. 5: Simulation results for Phi(+) across different areas of knowledge.
figure 5

Results from 19 simulations inwhich the highest weight for each of the 18 indicators was alternated. Panels represent units grouped by area of knowledge: a Humanities, b Exact Sciences, c Biological Sciences, d Engineering, and e Interdisciplinary Areas.

Full size image
Fig. 6: Results for the 18 indicators for two interdisciplinary units were analysed in the PROMETHEE application.
figure 6

Negative blue columns indicate the need to improve that specific indicator (e.g., a inclusion of black and mixed-race and b exchange students). Positive blue columns indicate sufficient positive effect of that indicator (e.g., a balance women and men and b international co-authorship).

Full size image

In the first simulation, academic units were identified by knowledge area but without assigning distinct weights to the 18 indicators (Table 4), whose actual data are available in statistical yearbooks and budget proposals considering the suggested dimensions in new model (Appendix 2). Figure 4 displays the overall outranking result of the units in PROMETHEE, with equal weights for the indicators.

Table 4 Indicators used in the simulations.
Full size table

By observing Fig. 4, it is possible to determine which units perform better in the set of indicators, considering the strengths (Phi+) and weaknesses (Phi-) resulting from PROMETHEE, assuming equal weights for the indicators (left side of Fig. 4). Additionally, Fig. 4 analyses units based on their respective areas of knowledge, including biological, engineering, humanities, exact sciences, and interdisciplinary areas.

Then, the units were separated by knowledge area, and 18 selected indicators were assigned different weights. The x-axis displays the indicator that received a weight of 10, while all others received a weight of 1. The variations in each unit demonstrate its behaviour based on the different weights assigned to each indicator. PROMETHEE’s Phi+ is emphasised to identify the best indicators for each unit by area.

In the humanities (Fig. 5a), assigning a higher weight to indicators such as events, number of students from public school, and the proportion of black and mixed-race individuals leads to better performance in the academic units. Conversely, in the exact sciences (Fig. 5b), higher weights assigned to productivity grants for faculties and researchers, FWCI, and foreigners in graduate programmes result in better performance. The results for the biological and engineering areas are more dispersed. However, better performance in biological sciences (Fig. 5c) is observed when a higher weight is assigned to the number of students in graduate courses with scholarship (IIC) and the number of graduated per entrants in graduate programmes. In engineering (Fig. 5d), FWCI and students on exchanging programmes stand out as important indicators. In interdisciplinary units (Fig. 5e), the indicator of event organisation stands out as a key factor. This analysis shows which indicators are most outstanding for each area and unit.

The application of PROMETHEE and ANP reinforces the theoretical framework by demonstrating how resource allocation can be optimised through a multidimensional approach that weights variables according to the contexts’ specificities. Thomas et al. (2020) argue that university evaluation models should incorporate a broad range of criteria, including academic productivity and social impact. The simulation results validate this perspective by integrating qualitative measures such as social engagement and diversity alongside traditional performance indicators.

In a second way for simulations, were conducted to analyse each unit’s results and visualise the performance for different indicators. The results of two of these simulations are presented below to demonstrate how indicators behave in different units and examine each improvement in detail. Figure 6a, b show the results of the simulation for two interdisciplinary academic units, considering equal weights for 18 indicators. Footnote 2

The blue columns below the bottom line indicate inferior performance compared to other academic units, highlighting potential areas that require improvement. The blue columns above the bottom line indicate superior performance compared to other faculties and institutes. By conducting this analysis, it is possible to guide each faculty’s or institute’s discretionary investment, revealing areas yet to be balanced. Figure 6a, b illustrate the two interdisciplinary institutes and faculties for the set of indicators within the model. The results shown in Fig. 6 indicate that the analysed faculty requires strategic investments in co-authorship with foreign authors and the number of scholarships for undergraduate students. This result of the institute in Fig. 6a indicates that incentives are necessary for international co-authorship and the proportion of students from public schools (blue columns below the bottom line).

Beyond the cost budget, additional funding can be allocated to promote improvements in specific indicators, considering the context of the university as a whole and the objectives of the strategic planning. Therefore, the budget must be aligned with the university’s planning objectives, considering the differences and particularities of each area. Once the units that require improvements in specific indicators are identified, strategic investments can be made.

The model implementation process comprises seven stages, systematically structured to optimise budget allocation within academic institutions. Initially, data collation (Step a) aggregates financial and performance data from disparate sources, thereby establishing a comprehensive dataset. Performance evaluation (Step b) utilises Data Envelopment Analysis (DEA) and Stochastic Frontier Analysis (SFA) to assess unit efficiency, thereby elucidating potential deficiencies. In the identification of limitations (Step c), the study delineates shortcomings inherent in the current model, thus necessitating a revised methodological framework. The definition of novel model premises (Step d) establishes broad performance indicators that are congruent with institutional objectives. Methodology selection (Step e) employs PROMETHEE and Analytic Network Process (ANP) to rank units based on a spectrum of criteria. Simulation and analysis (Step f) test a range of weight configurations, thereby ensuring data-informed refinements. Finally, resource allocation and strategic investment (Step g) direct resources efficiently to rectify weaknesses and align with institutional priorities. These stages are visually depicted in Fig. 7.

Fig. 7: Flowchart illustrating the seven stages of the resource allocation model implementation process.
figure 7

a Data Collation, b Performance Evaluation, c Identification of Current Model Limitations, d Definition of Novel Model Premises, e Methodology Selection, f Simulation and Analysis, and g Resource Allocation and Strategic Investment.

Full size image

Discussion and conclusions

The results validate the value of the proposed model, demonstrating its ability to comprehensively evaluate university performance by incorporating diverse dimensions such as societal engagement, inclusion, and sustainability. This supports the model’s relevance and effectiveness in guiding strategic decisions and enhancing university performance.

From the various objectives of research-intensive universities and organisational management, viewed from the changes that occurred from the republic of scholars of Bleiklie and Kogan (2007) to a stakeholder organisation (Musselin and Teixeira, 2013; Meek et al., 2010), universities have become increasingly complex environments and with multiple missions (Trakman, 2008). These missions tend to change over time incorporating new themes. In addition to traditional indicators, several emerging criteria, such as social impact, sustainability, and inclusion, are now essential to define a university of excellence. The evaluation of universities went beyond rankings and emphasised the importance of a comprehensive understanding of their objectives, policies, and practices to promote improvement. This requires considering the national, regional, and local context in which universities operate (Douglass, 2016; Agasisti and Berbegal-Mirabent, 2021; Chun and Sauder, 2023).

Several authors have provided evidence of the necessity and advantages of strategic planning in universities, demonstrating that those institutions that adhere to planning achieve positive outcomes (Fumasoli and Lepori, 2011; Nguyen and Van Gramberg, 2018; Gede and Huluka, 2023; Chen et al., 2024). Associating planning indicators with the budget is a way of stimulating the implementation of the strategic objectives of a university’s planning; this has been done for nearly two decades in several countries and universities and now with a tendency to expand.

Using DEA—Data Envelopment Analysis—and SFA—Stochastic Frontier Analysis— methods, we analysed the resource allocation case of one of the main universities, University of Campinas’s BQP. Our findings indicate that the programme effectively promoted the university’s objectives in the initial years after implementation, but quickly lost effect. The correlation between each BQP subprogramme and a single indicator, the presence of most units on the efficiency frontier, and the stabilisation of allocation highlights the need for new allocation criteria and models. When a stimulus occurs, budget allocation influences the collective movement in a uniform direction, with minimal adjustments. A point of stabilisation eventually occurs. Consequently, the pursuit of new indicators becomes necessary, and with enhancements in these indicators, they tend to gravitate towards stability in the future. Thus, fresh indicators should be considered in alignment with the objectives of quality and excellence (Douglass, 2016; Hinton, 2012; Elbawab, 2024).

Successful implementation in other institutions hinges on several critical factors. Firstly, access to comprehensive data on diverse performance indicators is paramount. Many universities may lack the infrastructure or data systems required to track the full spectrum of proposed indicators. Secondly, institutional leadership must demonstrate a firm commitment to adopting a multidimensional evaluation approach, which may necessitate a shift away from traditional, narrow performance metrics that some academic units may resist. Thirdly, academic units must receive adequate resources and training to engage effectively with the model and leverage its insights for informed decision-making.

Then, a practical resource allocation model is suggested in this work to guide academic units towards the university’s planning objectives strategically. This discussion is relevant because, since 1995, the university has not yet prioritised the change of quality indicators and has been unable to implement the necessary alterations. A plausible explanation is that the University has never analysed the subject as proposed in the current article, reproducing by inertia a rule that no longer makes sense. This new resource allocation model for research universities (which can encompass various types of resources beyond financial ones) was developed using MCDA (Multi-Criteria Decision Analysis) techniques and applied to the case of University of Campinas.

In addition to the proposed resource allocation model, several other factors contribute to achieving managerial efficiency and effectiveness. These factors include the following: developing careers that align with evolving goals (Knobel and Bernasconi, 2017; Bernasconi and Calderón, 2016); streamlining administrative processes to increase fluidity and speed while reducing time and motion (Pantel and Yakaboski, 2014); adopting professional communication practices (Trakman, 2008; Dearlove, 2002; Elbawab, 2024; Petersen, 2025); implementing modern teaching methods and providing managerial training for managers (Bernasconi, 2013; Pantel and Yakaboski, 2014); embracing new ways of thinking (Douglass, 2016; Javed et al., 2024; Chen et al., 2024).

As research universities must constantly strive to update their indicators, this paper makes a theoretical contribution by discussing research excellence concepts, elucidating the breadth of excellence and quality notions, and illustrating that quality indicators tend to converge when exposed to stimuli. Consequently, indicators should undergo periodic review, including the integration of new incentives. The significance lies in research excellence, underscoring how research universities, in their quest for excellence and quality, must sustain continuous efforts to revise their indicators and incentives.

From this perspective, several studies, analyses, mathematical tool development programmes, and application stages were conducted in this work, culminating in a table of indicators suitable for the case study. Through the analysis of the relevance of multicriteria decision-making methods for resource allocation, including ANP (Analytic Network Process) and PROMETHEE (Preference Ranking Organisation Method for Enrichment Evaluations), a model was proposed to identify which criteria can drive decisions in the context of research universities and prioritise resource allocation.

The novelty of the proposed model, beyond existing literature, stems from its multidimensional and adaptable framework for resource allocation in research universities. Unlike traditional models that rely on static and narrow performance metrics, this approach integrates strategic planning, social impact, sustainability, and inclusion, addressing contemporary challenges faced by universities. A key innovation is the dynamic weighting of indicators, achieved through the Analytic Network Process (ANP) and PROMETHEE methods, which allow indicator weights to be adjusted according to different areas of knowledge. This flexibility ensures that the model remains relevant across diverse academic disciplines. Moreover, it introduces a comprehensive approach to performance evaluation, incorporating not only research productivity but also internationalisation, financial sustainability, and community engagement. By aligning budget allocation with long-term strategic objectives, the model fosters institutional development in a way that traditional funding mechanisms do not.

Another distinguishing feature is its capacity for institutional customisation, enabling universities to prioritise indicators that reflect their specific missions and regional contexts. This marks a significant departure from one-size-fits-all budget allocation models, which often fail to accommodate the complexities of different academic environments. Additionally, the model emphasises proactive strategic alignment, allowing academic units to anticipate and adapt to evolving institutional priorities rather than merely responding to retrospective assessments.

Furthermore, the integration of mathematical decision-support tools ensures that funding decisions are both data-driven and context-sensitive. By addressing the shortcomings of existing budget allocation models, this research makes both a theoretical and practical contribution, offering a decision-making tool that research universities can implement to enhance institutional excellence and strategic coherence.

One of the limitations of this study is that the proposed allocation method is not intended to drive all the objectives of universities solely but to be one element of a set of actions that can guide universities in this direction (Fumasoli and Lepori, 2011; Nguyen and Van Gramberg, 2018; Gede and Huluka, 2023; Javed et al., 2024; Chen et al., 2024). Furthermore, the university must have the proposed indicators’ data organised and available by the academic unit to apply the complete model. Notably, the indicators proposed for model application are a basic, non-exhaustive list, allowing different units to include their indicators based on their areas of knowledge. Finally, the proposed methodology allows for future discussion and expansion of the set of indicators to meet the characteristics of the units, the incorporation of quality indicators, and periodic updating of indicators as the university’s strategic objectives evolve. The proposed model represents an advancement in how to distribute the budget. It moves away from limited indicators and aims at allocating resources to develop competencies, activities, inputs, and outputs that the units require. In the case here analysed, our approach considers the planning of the University of Campinas and seeks to promote the joint evolution of teaching, research, and extension.

The findings further support the theoretical frameworks proposed by Morgan Jones et al. (2022) and Moed (2020), which advocate for a holistic, flexible approach to university performance evaluation. By incorporating indicators related to social impact, inclusion, and sustainability, the proposed model illustrates how modern performance metrics can capture universities’ broader responsibilities. This aligns with recent literature calling for a more comprehensive assessment of academic performance beyond traditional productivity metrics (Oancea, 2019; Joly and Matt, 2017).

The model allows rewarding units in their respective areas, identifying which points need improvement in each unit, and directing efforts accordingly. Thus, a practical resource allocation model was proposed to consider the complexities and fronts involved, which can strategically guide the components of the University of Campinas towards the university’s objectives.

Future research directions

Future research may test the model’s adaptability and relevance across diverse university contexts, demonstrating its effectiveness in public and private institutions worldwide. Comparative studies will highlight its versatility by analysing how different university systems successfully implement it, reinforcing its universal applicability and transformative potential.

Additionally, longitudinal studies are recommended to assess the model’s long-term effects. Tracking its implementation over extended periods would offer a comprehensive analysis of its influence on academic units and institutional performance. This research would serve to determine whether the model fosters continuous improvement and facilitates sustainable development within universities. Long-term analyses could also evaluate how alterations in resource allocation impact academic outcomes, research productivity, and societal contributions.