Public R&D and Growth: A dynamic Panel Vector-Error-Correction Model Analysis for 14 OECD Countries
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsReview attached.
Comments for author File: 
Comments.pdf
The correction suggestions are attached.
Author Response
See uploaded doc (coverletter)
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThis paper addresses the significant issue of the impact of public R&D on economic growth. While the study provides valuable insights through its dynamic panel VECM approach, several areas require further clarification and revision to enhance its overall contribution.
The introduction section should not begin with a citation. A more engaging and compelling introduction to the study is necessary. Currently, the introduction resembles a literature review and should be completely revised.
The six items listed on page 2 are unnecessary in their current format. The authors should incorporate these items into sentence structure for better readability.
The data section should include a list of the countries studied.
Line 287: Variables with unit roots result in spurious regression rather than unstable regression.
Lines 293 & 301: The usage of three dots in these lines is unclear and requires an explanation.
Before applying the CIPS unit root test, the paper should test for cross-sectional dependence.
The paper employs multiple cointegration tests but needs to justify the selection of each test, especially given the presence of cross-sectional dependence. The rationale for using the Bai & Ng PANIC test as a "residual-based cointegration test" should be more thoroughly explained.
The description of the VECM model includes fixed effects and slope homogeneity, but the paper needs to further explain the assumptions underlying this specification and their relation to Park's (1998) theoretical model. The discussion should also address the potential limitations of slope homogeneity and the chosen method for handling cross-sectional dependence.
The shock analysis on public R&D is presented descriptively. The paper should interpret these results more explicitly, discussing their economic implications for policy decisions. Additionally, it should compare the results obtained from different VECM specifications (Model 1, Model 2, Model 3) to assess the robustness of the findings.
By addressing these weaknesses and implementing the suggested revisions, the paper can significantly improve its clarity, depth, and overall contribution to the understanding of the relationship between public R&D and economic growth.
Author Response
See uploaded document (coverletter)
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe topic of this manuscript is still very meaningful. However, the effects of public R&D on growth are undisputed and have been confirmed early on, while which have not yet been widely accepted as a quantitative analysis conclusion. This manuscript focuses on the improvement and application of the original econometric model, rather than discussing it from the most basic production function. Furthermore, it is recommended that the authors clearly outline all relevant variables.
The main problems are as follows.
1. The standardization of the manuscript is very poor! The keywords are actually mixed with the abstract!
2. There are too many footnotes in the manuscript. Suggest retaining only some necessary information.
3. Why have not the effects of public R&D on economic growth been explained from the perspective of production functions?
4. Suggest adding a table in “3. Data” to display the basic statistics of the main variables.
5. Some of the Notes in Table 1 are unnecessary, which should appear in “3. Data”, such as “Sample 1963-2017. 14 countries.”.
6. The drawing of Figure 1 and 2 on line 657 and 660 is not very good. I really can't see any other curves except for the blue one. A similar problem also appears in Figure 3 and 4.
7. “7. Summary and Conclusion” should be “7. Conclusions”.
8. Why did the authors not focus on analyzing the heterogeneity of domestic public R&D in the 14 OECD countries?
9. Domestic private R&D only appears twice in the text! Suggest increasing its analysis.
Comments on the Quality of English LanguageModerate editing of English language required
Author Response
See uploaded document (coverletter)
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsThe submitted manuscript addresses the controversial issue of the direct and indirect effects of public R&D on economic growth. The first section "Introduction" begins abruptly with a literature review. In this section the authors must introduce the readerint the research question and explain why they are focusing not ongross domestic savings in general but rather on the public R&D spendings. In lines 39-40 the author mentions six innovations. These are methodological extensions that have to be explained in detail only in the methodology section, not in the introduction!
The paper is fully free of theoretical framework section. This could be however accepted for empirical papers. I strongly recommend refering to Feder (1982) where the authors try to assess empirically the effects of exports on economic growth, where they do not just apply an aggregate production function but rather a multisector approach. In addition refer also to "Oil Abundance and Economic Growth" written in 2015 and published by Logos Verlag in Berlin. In this book, the authors use Feder (1982) as a theoretical framework from which they are deriving their econometric specification. At least briefly refer to the respective monograph.
Rename section 4 to Methodology. "Econometric methods" as a title is not usual. The authors must also refer to the methodological limitations in this section. The employed linear VECM approach does not assess the nonlinear effects. Their estimations are based on ARDL. Panel NARDL is a more advanced approach. This limitation has to be mentioned. Hence, refer to Shin et al. (2014) where NARDL has been proposed and further sources that account for the advantages of NARDL
https://doi.org/10.1017/jwe.2021.32
https://doi.org/10.1007/s13563-019-00202-6
Comments on the Quality of English LanguageThere are grave issues in this paper that have to be fixed. Otherwise, the paper cannot be published.
Author Response
See uploaded document (coverletter)Author Response File: Author Response.pdf
Reviewer 5 Report
Comments and Suggestions for AuthorsThis paper is well written and is publishable. The research methods used by the authors are appropriate. Results are reasonable and make sense. However, I have some comments for authors to improve the manuscript.
· Totally, there are 38 OECD countries currently in the world. But the authors only investigated 14 OECD countries. I believe that the authors need to provide some reasons to explain why you only chose 14 countries instead of all 38 countries and why these 14 countries were selected for this research (why not others). Why couldn’t you investigate all 38 OECD countries? Your explanations can help readers to better understand your main goal of doing this research.
· What are these 14 OECD countries? I believe that the authors should write down these 14 countries’ names in the third section – Data.
· The authors said, “More recent years in OECD MSTI are incomplete even for our 14 countries ….” What is the information that is incomplete? I believe that the authors need to tell readers what that information is and why it is incomplete.
· The authors said, “For technical change we use the data from Ziesemer (2023a) for labor-augmented technical change ….” Why couldn’t the authors get the updated data by yourselves instead of using past study’s data? The data used in the past study published in 2023 need to be updated.
· The authors said, “… using more recent sources that extend the time period to 1963–2017.” The authors need to update the data to cover the impact from the period of the Covid-19 pandemic (2020–2022) and the year of 2023. That is, the authors need to add six more years (2018–2023) to the dataset in addition to 1963–2017 to reflect the current economic situation.
· I believe that all OECD countries are not identical. Every country must have their specific culture, economic system, education system, political background, etc. Therefore, I suggest that the authors divide these 14 OECD countries (or more countries) into three or four categories to see how different characteristics lead to different results. The authors then can further discuss those differences, which can make the paper more valuable.
Author Response
See uploaded document (coverletter)
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
Comments and Suggestions for AuthorsThe authors have revised this manuscript according to relevant reviews.
The format standardization of the manuscript is still relatively poor.
Comments on the Quality of English LanguageMinor editing of English language required
Author Response
Please see attachment. Thanks for your comments.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsThe authors have addressed some of my comments. There are, however, no significant improvements. But, rather deterioration.
There are still grave issues in this paper. And, it starts from the very first sentence: "GDP per capita growth is driven mainly by technical change, although development 17 economics adds some other effects (Perkins et al. 2013)." GDP per capita growth is driven in the first lne by capital accumulation, and beginning from the steady state level, total factor productivity drives economic growth. The authors have to elaborate on this. In the transition phase, capital accumulation is the central driver. If you think that the countries of analysis are in the steady state in the respective time frame than you have to justify this and refer to Solow (1954). Against the backdrop of the time frame beginning in 1963 we can definitely say that most European countries were still in the transition phase and the assumption of a steady state is completely false.
Under such conditions, using linear models is a confinement. The author(s) have to account for this. The elaticities of the R&D investments cannot be assumed to be constant over the analyzed time frame!
Another issue is the introductory section that looks like a literature review. Please adjust the introductry part to the guidlines. Give a general introduction not the doscourse like you do it for instance here: "Erken et al. (2018) and Herzer (2022) emphasize ‘ambiguous results in the literature’ on public R&D and growth. Van Elk et al. (2015, 2019) analyzed the rate of return for public R&D for a panel of OECD countries."
To increase the readibility divide the aragraphs. The paragraph length of the first paragraph in the introduction is almost one full page. It is better to have 3 paragraphs for a page.
Furthermore, there is no problem statement and explanation on what is the contribution of this work to the literature, where is the research gap. This has to be made clear in the introductory section. Listing findings in the introduction is unusual. Avoid this.
The response related to the work of the WB veteran Gershon Feder (1982) shows that you are not familiar with this macroeconomic work. You assess the effects of the R&D investments on growth. This is in line with what Feder is also doing following the works of Balassa and others woth a focus on exports. Just read it and reassess your comment.
In section 2.4 the author states that she/he uses two approaches (lines 173-174). These are Cointegration analysis and VECM. This is not fully correct because cointegration test is the first step and the VECM is the second step.
Lines 183-187: ". It allows us to run simulations of shock effects going through all six equations for growth of en dogenous GDP, productivity, and the four R&D variables mentioned above." -- Explain which six equations are meant! After this sentence the author reports the findings from panel VECM. This is really confusing and shaows that the authors have to streamline and proofread the paper again.
The authors do not present a cross-section dependence test. They operate as if the probability of cross-section dependence is there. This leads to econometric redundancies. It is also not clear, why the authors do not use time dummy to account for joint shocks.
Libe 508: The following sentence is a complete nonsence: "Table 2 and Table A.1 both suggest that the problem of unit roots in the data cannot be ignored and thinking about cointegration is necessary"
Minor Comments:
Line 230: "We use data for 14 OECD countries for which we have relatively long data series: 230 AUT, BEL, DEU, DNK, ESP, FIN, GBR, ITA, JPN, NLD, NOR, PRT, SWE, USA." - This sentence is a n go for a reputable outlet! Put the countries in the dataset in a seperate table or in the footnotes.
Briefly refer to Sato (1963) for the incorporation of the ECMadjustment speed after fiscal (R&D spending) shock.
Comments on the Quality of English Language
English is sufficient.
Author Response
Please see attachment. Thanks for your comments.
Author Response File: Author Response.pdf
Reviewer 5 Report
Comments and Suggestions for AuthorsI read the author’s revision and response to my comments. The author mostly followed my comments to revise the paper. I am not very satisfied with the revision, but I can accept. In addition, Before the paper is accepted for publication, I request that the author should show each country’s whole name. For example, in the first paragraph of the data section, “AUT, BEL, DNK …” The authors should display Austria, Belgium, Denmark …
Author Response
Please see attachment. Thanks for your comments.
Author Response File: Author Response.pdf
