1. Introduction
Over twenty years of scholarship on the neoliberalism of higher education has captured its features, such as the corporate university, the entrepreneurial university, and the neoliberal university [
1,
2]. In neoliberal contexts, funding allocations for higher education have typically focused on competitive mechanisms. While higher education is funded directly by the state, it is usually seen as serving public goods, such as reducing inequality and increasing social mobility. Therefore, public goods-related policy initiatives seek to reframe higher education as interrelated with the well-being of society [
3,
4,
5]. Several governments have adopted the format of a national strategy or development plan by setting out national objectives for better alignment with higher education institutes, for instance, in Ireland, the Netherlands, Finland, and New Zealand [
6]. On neoliberal campuses, competition and evaluation are overemphasized. This phenomenon has received numerous criticisms [
1,
7,
8,
9,
10]. Considering the contemporary context of higher education, this study may provide a better understanding of the transformation from neoliberalism to public goods in the pursuit of sustainable higher education.
In addition, Eryaman and Schneider argue that research associations can promote the use of research to service public goods [
11]. Some nationwide and international associations have focused on this issue; for example, the Australian Association for Research in Education (AARE) identifies its vision of enhancing public goods by promoting, supporting, and improving research [
12]; the American Educational Research Association (AERA) recognizes the promotion of research to serve public goods as the fundamental responsibility of the association [
13]. Based on this, transferring public resources into higher education properly for public goods purposes has become a pertinent issue in the research community. Surprisingly, much less attention has been paid to verifying the effect of the allocation of resources through public goods in higher education settings. This study provides an alternative means to detect specific funding allocations for public goods purposes, which will provide a deeper understanding of this issue.
The funding policies in higher education are varied; for example, the UK has a different approach to that adopted in Germany, Italy, and the United States. Even the similar higher education systems in Japan and South Korea have specific considerations and differences. Taiwan is part of a different socio-political constellation, based on and driven by different value sets than neoliberal societies. Taking Taiwan’s Higher Education Sprout Project (HESP) as an example, this study explores how far the specific funding allocation can be transferred with the public goods implemented in higher education. The HESP (from 2018 to 2022, stage I) was established in 2017; it intends to transfer public goods as a policy-driven tool for sustainable institutes in higher education [
14,
15]. It represents a different direction for higher education that is aligned with sustainable policy. In this study, we argue that higher education for public goods purposes should consider balancing their teaching and research, caring for disadvantaged students to increase social mobility, and balancing global competition and local needs in terms of fulfilling social responsibility. This implies that higher education can be expected to achieve sustainable development under the specific funding allocation mechanism.
This study assumes that transforming public goods through special funding can play a critical role in neoliberal higher education, regardless of the sector. If the transformation model, input, transform, and outcome (ITO), works well, this suggests that public good initiatives can be implemented in higher education to achieve sustainable development. Therefore, this study focuses on specific policy implementation and uses the case study as an example to discover new knowledge in this field. If this approach is successful, the findings may encourage higher education institutes to commit to expanding learning opportunities for all, such as UNESCO’s SDG proposal for Education 2030. The purposes of this study include discovering the influential factors that exist in the funding allocation scheme and detecting the differences in funding allocation between systems and sectors, examining the effect of public goods transformation, and, finally, proposing better public goods strategies for higher education to achieve sustainable development. With these purposes in mind, the research questions to answer are as follows:
What are the influential factors in funding allocation in the HESP?
Did the funding allocation in HESP eliminate the diversity between the system and sector for public goods purposes?
Did a significant effect of the public goods transformation influence the target higher education system?
Can we set better strategies by way of specific funding allocations towards public goods in higher education for sustainable development?
This paper includes the following stages: First, we review funding allocation theories, the meanings of public goods, and their transformation logic in higher education. We utilize the policy initiative of HESP as an example of seeking public goods. Second,
Section 3 addresses the data collection and statistical processes conducted to verify the logic behind funding allocation. Third, we examine the effect of funding allocations in HESP using different types of higher education institutes and their structural relationships. Fourth,
Section 5 focuses on what challenges are confronted in the higher education system when public goods policy is intended to be implemented. Finally, conclusions are drawn and suggestions are provided for higher education.
3. Methodology
In this section, we address the research framework, logic of variable selection, sampling, data collection, statistical analysis, and verification of measure constructs. To explore the effect of HESP (2018–2022), this study employs a mixed method to examine funding allocation for public goods transformation.
3.1. Research Framework
The research framework is presented in
Figure 1. This framework displays how the funding is allocated for different institutes and its proportions for specific purposes. We consider the “system”, which refers to the two different tracks of institutes in terms of university and technological university systems in the target higher education; “sector” refers to the 157 public and private institutes in total. The models are considered by testing the funding effect of HESP and its influential factors with different approaches.
3.2. Design of Testing Model
This study designed an input, transformation, and outcome (ITO) model to explore the effect of funding on the quality of teaching, research, and public goods transformation in higher education. Different allocation of funding in HESP is considered as an input dimension. Students and faculty are considered as the major players in the transformation process. Since the teaching effect is not easily examined directly, we considered the expected outcomes to be academic efficiency in terms academic performance, internationalization, and participation in university social responsibility (USR). We assume that the simple model is the effect of input variables on expected outcomes, while the mediation effect might exert a significant impact. If this is the case that the mediation effect is significant, the impact of the transformation process might have a larger contribution in this model. Based on the research design, the major variables in the model are selected as follows:
3.2.1. Input Funding (I) Variables
Input funding refers to the funding in HESP for teaching, research, and public good purposes. The variables in the input funding include funding for HESP, funding per student, and funding for teaching.
“Funding_in_HESP” refers to the funds for 157 institutes in the target country. The total amount is TWD 15.34 billion (excluding the specific funding for selected research centers and the funding supported by the Ministry of Science and Technology).
“Funding_per_student” refers to the number of funds for students, which is the fund in each institute divided by the number of undergraduate students.
“Funding_for_teaching” refers to the fund for teaching purposes only. The calculation considered the fund for teaching divided by the number of undergraduate students in each institute.
3.2.2. Transformation Process (T) Variables
The transformation process refers to human resources and the mechanisms of transformation. The related variables in the transformation process include full-time faculty, international faculty, graduate students, and undergraduate students.
“Full_time_faculty” refers to the full-time faculty that the institute hired.
“International_faculty” refers to the full-time international faculty that the institute hired.
“Graduate” refers to the number of graduate students enrolled in the institution.
“Undergraduate” refers to the number of undergraduate students enrolled in the institute.
3.2.3. Expected Outcomes (O) Variables
This study defines the expected outcomes variables as academic performance, Scopus rank, number of USR projects, USR rank, and international students in each institute.
“Academic_performance” refers to the total number of journal articles for each institute in the Scopus database from 2018 to 2022. We assumed that the articles are related to research and might promote social well-being or solving global and local issues. The academic performance also indicates how institutes face global competition and global issues.
“Scopus_Rank” refers to the number of articles classified into four groups (Q1 to Q4) for the selected universities.
“USRNum” refers to the number of USR projects.
“USR_ranking” refers to the projects implementing social responsibility to fulfill local needs. This variable has been transformed, on a ranking basis, to compare the institute’s engagement in social responsibility. The USR_ranking was weighted by USR projects and funding.
“International_students” refers to the number of international students enrolled in the institute, representing a global competition indicator.
Figure 2 demonstrates the latent variables and their measurement constructs with specific indicators in this study. We proposed a structure of the input, transformation, and output process for further testing.
3.3. Sampling and Data Collection
The two most common methods of sampling are probability sampling and non-probability sampling. This study focuses on non-probability sampling techniques to collect institutional data. In this study, we conducted purposive sampling as an effective method for research in conditions where there is a confined target, for example, a specific period and a specific funding allocation. The data of institutes were collected from the Ministry of Education in Taiwan and the Scopus database (2018–2022) based on the targeted higher education institutes.
This study considered the students, faculty, and funding data of the 157 higher education institutes in Taiwan. The number of undergraduate and graduate students, international students, and faculty members was collected from the database of the Ministry of Education, Taiwan. Among these institutes, 50 institutes (31.85%) belonged to the public sector, and 107 (68.15%) belonged to the private sector. The university system consists of 71 institutes (45.22%), while 86 institutes (54.78%) are classified under the technological system. The “Full_time_faculty” ranges from 9 to 2045, and the range of “Undergraduate” is from 63 to 23,526 in 2022. The funding and USR data are based on a document published by the Ministry of Education in 2018 [
15]. The institutional “Academic_performance” data are based on the Scopus database, and the data were collected from 2018 to 2022. Most of the data are secondary data. We integrated and transformed the data to fit the requirements of quantitative approaches. In PLS-SEM, 116 samples (73.89%) were selected to fit so that there is no outlier of “Academic_performance”. Therefore, we exclude four universities for global Taiwan and the institutes for which there is no journal paper information available. Finally, there are 78 private institutes (49.68%) that fit the selected criteria. Based on Hair et al.’s suggestion, a minimum sample size of 52 for PLS-SEM that has statistical power of 80% in the study is used [
52]. The samples in this study fit the minimum requirement.
3.4. Statistical Analysis
In this study, the Statistical Package for the Social Sciences (SPSS 26), Minitab 20, and partial least square structure equation modeling (PLS-SEM 3.0) were used to analyze the data-transforming-related evidence to support our arguments. PLS-SEM was conducted to consider the holistic structure of our testing model, while the intervariable influences were determined by regression and logistic regression models. The procedures for statistical analysis are as follows:
First, the influential factors in the transformation process are checked. Stepwise regression models with specific variables are used to test the effect of academic performance in the institutes [
53]. This study compared two different models to check the logic of funding with regression analyses. One includes all the possible institutes to interpret the effect of funding on the academic efficiency in the model. The other excludes the selected top four universities to determine which variables critically influence funding allocation without considering academic excellence. The dependent variables are “Funding_in_HESP” (unit: TWD 10,000) and “Funding_per_student”. The related independent variables are selected by the stepwise method to build fitted regression models to interpret.
Second, logistic regression is conducted using Minitab to determine the effects of funding allocation in HESP on the sector and different higher education tracks. The sector and track of universities are categorically coded variables and dependent variables in the logistic models. The logistic regression estimates the probability of an event occurring, such as voting or not voting, based on a given dataset of independent variables. Since the outcome is a probability, the dependent variable is likely between 0 and 1. In the logistic regression, a logit transformation is applied to the odds—that is, the probability of success divided by the probability of failure. The odds ratio (OR) was calculated to reflect the effect of funding allocation in HESP on the sector and system of the current higher education institutes. The OR was calculated according to the following formula with conditions A and B [
54]:
We also considered the stepwise method with more complicated models in the logistical regression. In the significance tests, the critical value was set to α = 0.05.
Third, PLS-SEM was used to verify the effect of funding allocation in the HESP. We selected funding for the institute, funding per student, and funding for teaching as formats of funding allocation in HESP. Full-time faculties, international faculties, undergraduate students, and graduate students are the human resource variables that represent the transformation process. Academic performance, Scopus_Rank, USRNum, USR_ranking, and international students are the expected outcome variables. Typically, PLS-SEM is employed to model the relationship between the measured and latent variables or between multiple latent variables. Since multiple regression is restricted to examining a single relationship at a time, PLS-SEM can estimate a series of interrelated and dependent relationships simultaneously. This technique enables researchers to quickly set up and reliably test hypothetical relationships among theoretical constructs and those between the constructs and their observed indicators. PLS-SEM is more effective than multiple regressions in parsimonious model testing. It is employed to find the best-fit model [
55]. Moreover, PLS-SEM is more powerful than covariate-based structure equation modeling (CB-SEM), and it can be applied on non-normal data and a relatively small sample size [
56,
57]. PLS-SEM is also recommended over CB-SEM when the model is complex, and it aims to test the theoretical framework [
58,
59]. Both types of SEM also have potential biases. However, PLS-SEM tends to wield greater power to minimize the biases [
59,
60]. This study considered that PLS-SEM can deal with non-parameter data, construct linear structures in a testing model quickly, and verify the testing constructs. Moreover, PLS-SEM provides a set of goodness-of-fit indices to support the model’s robustness, for example, quality criteria, reliability (internal consistency), discriminant validity, and bootstrapping.
3.5. Verification of ITO Measure Construct
This study considered the overall model fit in PLS-SEM using the following goodness-of-fit indices: Quality criteria, constructs, effects, r
2, discriminant validity, and residuals [
56,
58,
61,
62]. Since differences between sectors might exist, we also verified the effect in the target higher education with PLS-SEM. Based on previous studies’ suggestions, we verified the measure construct with Cronbach alpha (>0.7), average extracted variance (AVE > 0.5), composite reliability (CR > 0.7), heterotrait–monotrait (HTMT) ratio (<0.85 or <0.9), critical path, and its coefficient in the PLS-SEM [
52,
63,
64]. Cronbach alpha and CR were considered to determine internal consistency. AVE was used to analyze convergent validity. HTMT ratio is an estimation of the correlation between the constructs and it is a criterion to assess discriminant validity [
64]. Kline suggests a threshold of 0.85 or less in HTMT [
65], while Teo et al. recommended a liberal threshold of 0.90 or less [
66].
Finally, following Shrout and Bolger’s suggestion [
67], the bootstrap method was used to estimate the mediation effect in this study. We resampled 2000 for bootstrapping to demonstrate that the proposed model is robust. When Z > 1.96 (Z = point estimate/standardized error), this implies that there is a mediation effect among the latent variables [
68,
69]. When the mediation effect exists, we confirm the effect in the transformation process.
5. Discussion
In Taiwan, previously enhanced quality and introduced teaching excellence programs are based on the competitive mechanism [
51]. Adverse effects have been reported, for example, the overemphasis on evaluation and the necessity for accountability in a short period of time [
70]. Moreover, studies from the perspectives of students and teachers indicate that universities receiving teaching excellence program grants failed to meet their expectations [
71,
72]. This is why the HESP was initiated. Can public goods work well in higher education with a series of policy-driven reforms in neoliberal contexts? This is a critical point of public goods transformation for sustainable higher education. In the beginning, the HESP considered targeting the quality of higher education institutes, balancing institutional excellence, and improving the quality of teaching for disadvantaged students. Specific funds from HESP are offered to all higher education institutes instead of a competition scheme. In addition, this study found that the expected outcomes are academic performance and international student recruitment, whereas the impact of USR is still limited. In PLS-SEM testing, the findings suggest that the initial funding provided by HESP can affect the expected outcomes through the transformation process. The mediation effect of the transformation process is significant in the proposed model. Compared with previous policy initiatives, the most significant change in the HESP is funding allocation for all institutes and focusing on implementing USR. USR consists of strengthening university–industry collaboration, fostering cooperation among universities and high schools, and nurturing talent required by local economies. In the long term, the influence of USR projects will increase in the higher education system. In this sense, HESP demonstrates that USR could be a crucial factor in a novel sustainable model.
In this study, we also raised two crucial questions: How wide is the gap in the funding allocations in HESP between the system and sector for public good purposes? What are the influential factors for funding allocations in the HESP? HESP encouraged higher education institutes to promote teaching innovation by enhancing learning effectiveness and teaching quality to reduce inequality. Based on the effect of funding allocations, this study found that some issues are emerging in the HESP. First, the HESP aims to secure students’ equal rights to access good-quality and diverse higher education systems. Suppose equal rights to access higher education reflects no significant difference in the funding allocations for institutes. While the funding scheme reveals that the institutional scale has been considered in the funding allocations, there is a small gap between universities and private technological universities and colleges. Fortunately, the results of PLS-SEM confirm that the transformation process works well in both sectors. Second, this study found that the funding of the HESP for each institute is based on the number of articles in Scopus due to their high association in the testing model. The government encourages higher education to propose institutional projects with unique characteristics for their sustainable development. At the same time, our results reveal that there is a similar culture on campuses where encouragement for article production persists. Therefore, balancing teaching and research should be considered in the next stage of HESP.
In a global context, higher education policies have typically been shown to provide incentives for universities to develop or strengthen their capacity for the academic profession and performance in neoliberal times [
2,
7]. For example, Codd’s and Burton-Jones’s arguments reflect a similar phenomenon in higher education [
73,
74]. Various funding-centered studies have focused on global competition discussions in neoliberal contexts [
10,
39,
41]. In comparison, various studies indicate that the concept of public goods might play a significant role in higher education [
11,
34,
36]. Like the EU agenda for higher education and the new global Education 2030 [
4,
5], Ireland’s National Strategy for Higher Education to 2030 and the Dutch National Research Agenda also provide ambitious targets for public goods in higher education. This indicates a possible transition within the neoliberal regime from competition-oriented to public-good-oriented systems.
Can significant policy initiatives transform public goods in neoliberal higher education settings? This study provides an empirical example (an ITO model) to evaluate the core values of public goods and their practices in higher education settings. The findings suggest that when higher education is considered from a public goods perspective, the competitive funding scheme should consider the policy’s intention and the effect of implementation. Even though the policy intention is evident in this case study, change still needs to be faster and more predictable in neoliberal higher education contexts. This study confirmed that the common good is a collective decision that involves the state, the market, and civil society [
30]. Since it is impossible to exclude any individual from benefiting from the good [
26,
27], higher education policies for public goods intervention may need adequate resources for long-term support. With appropriate funding for public goods purposes, higher education can find ways to respond to the challenges of local and global issues. This study found that current policy intentions and short-term funding support did not match well.
Taking HESP as an example, the ITO model may provide a more holistic perspective to reflect the issues in neoliberal higher education settings, regardless of the public and private sectors. With higher education institutes, the effectiveness of education, research, and innovation can appropriately connect to societies. As stated in previous discussions, the private sector does not usually provide pure public goods; therefore, pure public goods in higher education are a minor phenomenon [
48]. In this study, we demonstrate that the effects of specific funding for public goods are significant regardless of the sectors in higher education. The example of HESP may provide a more profound understanding of funding allocations for public goods in neoliberal times. Even though the private sector received limited public funds in the HESP, funding-driven policy supported all private institutes to achieve public goods in this case study.
In general, performance funding is based on an input/output model of services, where services are financed by government agencies in terms of output indicators. As higher education moved into a globally competitive era, questions arose concerning putting public goods schemes to work in a neoliberal context. What will be the effect of public goods perspectives on contemporary higher education? After reviewing the relevant literature, this study observed changes. For example, Tian and Liu’s study indicated that public or common goods also triggered discussions concerning higher education in China [
37]. Huang and Horiuchi addressed the public goods of internationalizing higher education in Japan. Despite the acceptance of the concept of public goods, changes and reforms in the Japanese system have been dominated by demands from business and industry [
49]. Moreover, many European countries have implemented some form of performance-based funding in higher education. For example, implementing research-performance-based funding (RPBF) systems aims to improve research cultures and facilitate institutional changes that can help increase research performance [
75]. Many EU countries have introduced, are introducing, or are considering introducing such systems, whereas the consideration of the implementation of public goods, tuition, and fees has traditionally been low in Europe, reflecting the view that higher education is a public good [
76]. There are alternative funding schemes to fit various performance purposes in European countries.
6. Limitations and Suggestions for Future Studies
This study may confirm that only when the balance between equity and excellence is achieved can a sustainable higher education system be expected. Higher education is the most diverse system in the world. Within this diverse system, the equity issue has been raised in many countries. For example, Zerquera and Ziskin’s study indicated that performance-based funding requirements interact with the public-serving mission of urban-serving research universities (USRUs) in the USA and can deepen stratification across a differentiated system [
77]. Therefore, the ITO model has its limitation to fit all the diverse systems.
Moreover, considering disadvantaged groups, various studies have focused on how performance-based funding impacts marginalized students [
78,
79,
80,
81], with findings across these studies essentially pointing to adverse effects on access for underrepresented students. Unfortunately, this study did not find significant evidence to support the argument that HESP positively affects underrepresented students. Therefore, policymakers, institutes, and researchers should work towards synergistic interactions to deepen their understanding and vision through transforming higher education for a better society. In the next transformation stage, it is essential that cooperation leads the program to success in higher education.
This study might be limited in its quantitative approach. There is great information in the context of practices that might be neglected in this study. Therefore, related qualitative approaches are alternative strategies that could be used to access and interpret other kinds of data.
For future studies, we encourage researchers to consider PLS-SEM in related policy studies to provide help information for policymakers. The reason is PLS-SEM can provide a non-parametric model to test cause–effect hypotheses in policy settings. We also suggest that local researchers follow up on the effect of stage II HESP from 2023 to 2027. For international researchers, the design of the study can be extended to tackle similar issues in other higher education settings. Since sustainable development covers social, economic, and environmental issues, the notion of public goods transformation is not limited to higher education settings only. Public goods transformation issues have emerged for different reasons and at different levels of organizations. Similar institutes can also think about the notions and models that can be developed in society in the future.
7. Conclusions
Our findings suggest that the government initiated the HESP and targeted the quality of higher education institutes to balance institutional excellence with caring for disadvantaged students’ learning. Regarding this core issue, policy managers need to engage in continuous discussion with partners to overcome the funding gaps for public goods purposes. Our study demonstrates that public goods have been transformed in higher education by reshaping what universities are expected to do in an uncertain future. This case study may provide a valuable reference when policy design considers both theories and practical issues by transforming public goods for sustainable higher education.
Moreover, this study focuses on the following concerns for higher education: First, reshaping institutional strategies for public goods and promoting strong institutional characteristics for substantive development in the future are necessary. Second, it is crucial to continue balancing academic excellence and quality teaching, and commitment to implementing innovative and quality teaching for disadvantaged groups should be the premier institutional strategy for most institutes. Third, higher education institutes should commit to achieving remarkable progress in expanding learning opportunities for all as part of the UN’s SDGs. With appropriate funding, higher education can find ways to respond to the challenges of local and global issues.
Since achieving sustainable higher education is a long-term goal, we know it needs many resources and partners to support it. Therefore, we hope that the original design and novel verified approaches in this case study provide helpful examples to explore similar issues in higher education settings.