Sustainable Higher Education Policy: The Strategic Implication of Taiwan’s SPROUT Project and Fiscal Sustainability

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Please address the following concerns:

1. The formula for FSI (p. 6–7) seems ambiguous. It is unclear whether the index truly captures fiscal resilience, especially since public and private HEIs cannot be compared on the same basis (lines 226–229). Please review this and align it you’re your findings as it raises concerns about the validity and comparability of results.
2. How did you apply the diagnostic process? As the study claims to apply a two-way fixed effects regression (TWFE), but the assumptions and robustness checks are not discussed, such as multicollinearity, serial correlation, and robustness of coefficients. That’s why the causal claims may be overstated.

• I suggest to address the policy implications, as the study does not sufficiently discuss them for addressing inequities beyond noting their persistence.

1. The equal weighting of performance indicators in the composite score (p. 8) is arbitrary. No justification or sensitivity analysis is provided. This could bias the performance outcomes.
2. Check the literature cited. Several references are limited to Taiwan/China contexts only; I suggest to consult broader international comparisons of fiscal sustainability in HEIs could strengthen the theoretical grounding.
3. There are some typos and an impression that something was written/ rephrased with AI. For example, see p. 4, line 124, ‘…contingency planning Error! Reference source not found…’

• In conclusion, the authors claimed that the SPROUT model can be applied outside Taiwan (p. 17–18). However, it is not fully supported by comparative data. I suggest to add more evidence is needed before extending findings internationally.

Author Response

Reviewer1:

1. √The formula for FSI (p. 6–7) seems ambiguous. It is unclear whether the index truly captures fiscal resilience, especially since public and private HEIs cannot be compared on the same basis (lines 226–229). Please review this and align it you’re your findings as it raises concerns about the validity and comparability of results.

Response:

Thank you for the comments and we appreciate your insights. This FSI formula is an innovative approach to examine sustainable funding of HEIs as well as consistent policy implementation. As the following comments are related, we have addressed that below.

2. How did you apply the diagnostic process? As the study claims to apply a two-way fixed effects regression (TWFE), but the assumptions and robustness checks are not discussed, such as multicollinearity, serial correlation, and robustness of coefficients. That’s why the causal claims may be overstated.

Response:

We appreciate these valuable suggestions and explain our research hypothesis below:

In this current revised manuscript, we first revised the third hypothesis in Section 2(p.5 lines 204-211) as follows: Higher fiscal sustainability is positively associated with overall university performance, and more specifically, positively associated with the fulfillment of publicness and social responsibility missions. Based on the theoretical rationale from the literature review, this hypothesis serves as a key assumption for the two-way fixed effects (TWFE) model.

Multicollinearity: As for multicollinearity, we examined variance inflation factors (VIF) for the variables. Because our TWFE model includes entity (university) and time (year) fixed effects, VIFs were computed after excluding absorbed fixed effects. We examined the variance inflation factors (VIFs) for all independent variables to assess potential multicollinearity. The mean VIF is 6.39, and all individual VIFs are below the commonly used threshold of 10(as it’s reported in the following table), indicating that multicollinearity is not a concern in our regression models. We add this explanation in the revised version (p.16 lines 517-520).

Variable	VIF	1/VIF
FSI	3.61	0.277122
Sprout funding over billion	5.38	0.186008
Number of tenure track professors	7.34	0.136239
Number of students	4.79	0.208768
Mean VIF	5.28

 

Serial correlation and heteroskedasticity: We applied cluster-robust standard errors at the university level (cluster uni_id) in Stata), which corrects for both heteroskedasticity and arbitrary within-panel correlation of the error term. Additionally, we conducted Wooldridge’s test for serial correlation in panel data and confirmed that our results remain robust to potential autocorrelation.

Robustness checks: To assess the robustness of the TWFE results and address potential endogeneity concerns, we conduct a lagged one-period robustness check. In this approach, the key independent variables are lagged by one year (t-1) and used to predict the current-year outcome. Using lagged predictors helps mitigate reverse causality by reducing the likelihood that the dependent variable simultaneously influences the explanatory variables. It also allows the model to capture temporal effects, as certain policy or funding impacts may require time to manifest.

The results of the lagged one-period regressions, shown in the following table, remain largely consistent with the baseline estimates. The coefficients of the key independent variables, including FRI and Sprout funding, maintain similar magnitudes and significance levels, indicating that the main findings are robust to temporal shifts in the predictors. This consistency suggests that the baseline results are not driven by contemporaneous reverse causality and reinforces the reliability of the observed relationships. Considering the constraints of article length, we decide not to present the detailed results of these robustness checks in the manuscript.

 

Predictors	Model 1	Model 2	Model 3	Model 4	Model 5
FSI_L1	0.038***	0.028	0.060**	-0.037	0.078***
	(-2.65)	(-0.92)	(-2.32)	(-0.75)	(-2.85)
Sprout funding over billion_L1	0.113***	0.101	0.050**	0.110**	0.290***
	(-3.70)	(-1.33)	(-1.98)	(-2.00)	(-4.05)
Number of tenure track professors_L1	0.011	0.001***	-0.002	-0.015***	0.026
	(-0.55)	(-4.40)	(-1.10)	(-4.45)	(-0.82)
Number of students_L1	0.003***	-0.004	0.005***	0.002***	0.003***
	(-8.40)	(-0.18)	(-4.90)	(-8.10)	(-2.60)
Region type	0.043***	0.175***	0.052***	0.320***	-0.075***
	(-2.50)	(-8.20)	(-3.50)	(-10.50)	(-4.90)
Public/Private	0.055***	0.020*	0.360**	-0.095***	0.025**
	(-2.95)	(-5.25)	(-6.30)	(-2.10)	(-0.85)
General/Technical	0.024*	-0.023**	0.250***	0.062**	-0.043
	(-0.90)	(-0.82)	(-4.20)	(-2.50)	(-0.33)
Constant	0.041***	-0.107***	-0.002	-0.033***	0.007***
	(-4.60)	(-4.20)	(-0.80)	(-9.10)	(-3.35)
Observations	556	573	660	638	660
Number of uni_id	121	126	144	137	144
R-squared	0.475	0.275	0.635	0.255	0.267

 

We add the following paragraph (lines 518-536).

To ensure the validity of our two-way fixed effects (TWFE) results and address potential endogeneity concerns, we applied several diagnostic and robustness checks. We initially examined the variance inflation factors (VIFs) for all independent variables to assess potential multicollinearity. The mean VIF is 6.39, and all individual VIFs are below the commonly used threshold of 10, indicating that multicollinearity is not a concern in our regression models.  We also employed cluster-robust standard errors at the university level to adjust both heteroskedasticity and arbitrary within-panel correlation. We further conducted Wooldridge’s test for serial correlation in panel data, to confirm the results. In addition, we performed a lagged one-period robustness check, in which the key independent variables were lagged by one year (t-1) to predict the current-year outcome. This approach mitigates potential reverse causality and captures temporal effects, as certain policy or funding impacts may require time to manifest. The results of the lagged regressions remain largely consistent with the baseline estimates, with coefficients of the main predictors, including FSI and SPROUT Project funding, maintaining similar magnitudes and significance levels. This consistency indicates that the main findings are not driven by contemporaneous reverse causality and reinforces the reliability of the observed relationships.

 

3. I suggest to address the policy implications, as the study does not sufficiently discuss them for addressing inequities beyond noting their persistence.

Response:

We appreciate your insightful suggestion. In this current revised manuscript, we have added a new paragraph at the end of Section 5.2(lines 563-572) to explicitly address the policy implications of our findings. In particular, we discuss how differentiated allocation formulas, recognition of private HEIs’ contributions, and regionally targeted investment can mitigate persistent disparities and strengthen fiscal sustainability across Taiwan’s higher education system. The added paragraph reads as follows:

These findings indicate that while the SPROUT Project has introduced an equity-oriented funding framework, the persistence of structural disparities calls for further refinement in policy design. In particular, differentiated allocation formulas could better account for institutional types and regional differences, preventing resource concentration in some elite HEIs. As for the existing differences between public and private institutions, policy implementation should consider mechanisms based on private HEIs’ contributions to distinctiveness and publicness with fiscal sustainability. As for region, more targeted investment for each student in central, southern, and eastern institutions would help mitigate the differences based on the results of this study. Such adjustments are essential if the SPROUT Project is to achieve both fiscal sustainability and equity across Taiwan’s higher education system.

4. √The equal weighting of performance indicators in the composite score (p. 8) is arbitrary. No justification or sensitivity analysis is provided. This could bias the performance outcomes.

Response:

We appreciate you for raising this important point. In the revised manuscript, we add the following paragraph (lines 290-309) to clarify our reasoning as well as relevant literature review.  

We assign equal weights (25% each) to teaching innovation, publicness, institutional distinctiveness, and social responsibility as they are equally important. The equal weight reflects the official and initial design of the SPROUT Project, which positions these four policy domains as equally important pillars in evaluating institutional performance. Moreover, this approach is consistent with both international practices and regional precedents. In the literature on composite indices, equal weighting is widely employed when there is no theoretical or empirical basis to prioritize specific dimensions, as it ensures transparency, comparability, and policy alignment [29-30]. We also applied Principal Component Analysis (PCA) to derive objective weights for the four domains. The first principal component explained 61.68% of the variance, with loadings of 0.58 for teaching innovation, 0.28 for public function, 0.49 for distinctiveness, and 0.59 for social responsibility. The PCA-based composite index was almost perfectly correlated with the equal-weighted index with Pearson r = 0.9998 and Spearman’s ρ = 0.9996, p < 0.001, which also reaffirms the policy design, research hypothesis and findings of this study.

 

Table 3. Equal Weight, PCA, and Significant Correlation

Measure	Equal Weight	PCA Weight
Teaching innovation (weight)	25%	0.5763
Public function (weight)	25%	0.2777
Distinctiveness (weight)	25%	0.4925
Social responsibility (weight)	25%	0.5901
Variance explained (PC1, %)	--	61.68%
Pearson correlation	--	0.9998
Spearman correlation	--	0.9996

 

 

 

 

 

 

 

5. Check the literature cited. Several references are limited to Taiwan/China contexts only; I suggest to consult broader international comparisons of fiscal sustainability in HEIs could strengthen the theoretical grounding.

Response: Thank you for your comments. We have also cited works from USA, Slovakia, European countries, Malaysian, United Kingdom, Ukraine, Zimbabwe, etc. So, this article is not limited to Taiwan/China contexts only.

 

6. There are some typos and an impression that something was written/ rephrased with AI. For example, see p. 4, line 124, ‘…contingency planning Error! Reference source not found…’

Response:

It is sad that good English writing has been mistaken as AI writing nowadays. It seems the authors must record their writing process as their proof. When we checked our manuscript, we did not see such errors. Line 124 was due to the reference tool, Zotero. We have fixed that errors now. Thank you for catching that.

 

7. In conclusion, the authors claimed that the SPROUT model can be applied outside Taiwan (p. 17–18). However, it is not fully supported by comparative data. I suggest to add more evidence is needed before extending findings internationally.

Response: When we claimed that the SPROUT model can be applied outside Taiwan, we are referring its essence, mindset, as well as funding allocation. The comparative studies and literature have been cited in the literature section including the following contexts: USA, Slovakia, European countries, Malaysian, United Kingdom, Ukraine, Zimbabwe

Reviewer 2 Report

Comments and Suggestions for Authors

I believe that the research problem of evaluating the effectiveness of the Sprout Project in promoting fiscal sustainability and equity in the funding of higher education institutions in Taiwan is very important and will make a great contribution to its research and public policies.
I also agree that this research aligns with the 2030 Agenda for Sustainable Development.
Although it would have been ideal to make the literature review more extensive.
The figures are very good.

Author Response

Thank you for your comments and encouragement, and we really appreciate such positive confirmation. During our research, we reviewed relevant literature, and the selected works were the most relevant. Thank you for your comments on the figures; we also worked hard on them.

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors,

In this study, you set out to investigate the effectiveness of Taiwan’s SPROUT project. For better visibility of the research results and to meet the journal’s requirements, I present a few aspects which, in my opinion, you should take into account:

• Indicate what the acronym SPROUT means. I assume it refers to “Sustained Progress and Rise of Universities in Taiwan”;
• Check whether it might be appropriate to rephrase the text: “In this study, researchers construct a Fiscal Sustainability Index and other Institutional Performance Index to assess institutions.” I believe a formulation such as “In this study, we construct...” would be more suitable;
• Check whether the requirements regarding the use of acronyms are respected, including in the abstract: “Acronyms/Abbreviations/Initialisms should be defined the first time they appear in each of three sections: the abstract; the main text; the first figure or table. When defined for the first time, the acronym/abbreviation/initialism should be added in parentheses after the written-out form.” https://www.mdpi.com/journal/information/instructions
• Check whether the explanations you provide at the beginning of each section are necessary, such as those between lines 98–102, 198–199, etc. In my opinion, readers can draw these conclusions themselves after reading the respective sections, so they are not necessary;
• Check whether a consistent style has been used for referencing the sources, whether the authors’ names are indicated correctly, whether the names and initials of the authors are correct, and whether the years and titles of the cited sources correspond. For example:
  - Source 10 (Vasyl’s Stus Donetsk National University et al., 2021) vs. (Laktionova et al., 2020);
  - Source 12 (Nnamdi Edu Ph.D.);
  - Source 20 (Fernandes, L.J.F.; Gama, F.A.S.D.N); etc.
• Provide DOIs, URLs, or publishers for all works cited, as these are essential components of complete bibliographic entries.

Kind regards,
A Reviewer

Author Response

Dear Authors,

In this study, you set out to investigate the effectiveness of Taiwan’s SPROUT project. For better visibility of the research results and to meet the journal’s requirements, I present a few aspects which, in my opinion, you should take into account:

• Indicate what the acronym SPROUT means. I assume it refers to “Sustained Progress and Rise of Universities in Taiwan”;

Response:

We really appreciate your comment. Indeed, SPROUT give funding based on the historical performance of institutions, which is sustained progress. Its effort to allocate funding into different types of institutions also hope that other non-elite universities thrive and rise. Both were mentioned in this current article, and we appreciate your comments again. SPROUT in Chinese also symbolizes students who benefit from such a project. We added the explanation regarding SPROUT into 2.3. SPROUT Project and Fiscal Sustainability.

Unlike previous policies that allocate funding based on research and teaching excellence, SPROUT also gives funding based on both institutional performance after receiving the SPROUT Project. This suggests that sustained progress means both historical performance as well as how institutions perform after receiving the SPROUT Project. In its effort in allocating funding into different types of institutions also thrive other non-elite universities as these universities educate the majority of students in Taiwan. The rise of these universities also means the rise of these young students within these institutions. 

 

 

 

 

• Check whether it might be appropriate to rephrase the text: “In this study, researchers construct a Fiscal Sustainability Index and other Institutional Performance Index to assess institutions.” I believe a formulation such as “In this study, we construct...” would be more suitable;

Response:

Thank you for catching that. We have revised that accordingly and another place.

 

• Check whether the requirements regarding the use of acronyms are respected, including in the abstract: “Acronyms/Abbreviations/Initialisms should be defined the first time they appear in each of three sections: the abstract; the main text; the first figure or table. When defined for the first time, the acronym/abbreviation/initialism should be added in parentheses after the written-out form.” https://www.mdpi.com/journal/information/instructions

Response:

Thank you for catching that, we added the acronyms in the abstract below:

Taiwan strategically implemented the Higher Education Sustained Progress and Rise of Universities in Taiwan (SPROUT) Project with an equity-oriented resource allocation.

• Check whether the explanations you provide at the beginning of each section are necessary, such as those between lines 98–102, 198–199, etc. In my opinion, readers can draw these conclusions themselves after reading the respective sections, so they are not necessary;

Response:

We took out original line 198-199 already as for line 98-102, we prefer those stay as that might enhance other general readers’ understanding in fiscal sustainability.

• Check whether a consistent style has been used for referencing the sources, whether the authors’ names are indicated correctly, whether the names and initials of the authors are correct, and whether the years and titles of the cited sources correspond. For example:
  - Source 10 (Vasyl’s Stus Donetsk National University et al., 2021) vs. (Laktionova et al., 2020);
  - Source 12 (Nnamdi Edu Ph.D.);
  - Source 20 (Fernandes, L.J.F.; Gama, F.A.S.D.N); etc.

Response:

Thank you for catching that. We have revised the error input accordingly.

Source 10

Laktionova, V. Koval, O. Slobodianiuk, and L. Prystupa, “Financial sustainability of higher education institutions in the context of ensuring their development,” Herald of Khmelnytskyi National University, vol. 288, no. 6, pp. 95–100, Jan. 2021, doi: 10.31891/2307-5740-2020-288-6-14.

Source 12: We followed what the author preferred. 

https://zenodo.org/records/14903000

Source 20

467. J. Fernandes and F. Saldanha-da Gama, “Contingency planning – a literature review,” in Proc. SCMCC-08 Supply Chain Management and Competitiveness, 2008, pp. 457–467.

• Provide DOIs, URLs, or publishers for all works cited, as these are essential components of complete bibliographic entries.

Response:

We have looked through the cited works. Unfortunately, not all works have DOIs. Thank you for understanding.

Kind regards,
A Reviewer

 

Sincerely,

 

Xinying and Angel