Analysis of Effects of COVID-19 Pandemic on Small- and Medium-Sized Enterprises (SMEs) in Rwanda Using Wood Firm-Level Data

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

After reading the entire article, I noticed the following aspects related to:

Abstract. The authors specify clearly what is the main objective of their research, the methodology used and the results obtained as a result of the research study.

Introduction. The authors present the general framework of the topic addressed and some of the implications of COVID-19 on SMEs from several countries. However, the authors do not structure this part of the Literature review on a certain geographical or continental area in order to make comparisons or to provide additional clarifications regarding the connection with the study undertaken by them. I suggest the authors to structure more clearly the implications of COVID-19 on SMEs on continental areas (for example Europe - Africa or America) to highlight certain results or implications of the epidemic with a greater emphasis on the wood business area, as the authors have pointed out in the main research objective.

I also suggest the authors to underline this impact of COVID-19 on the wood business right in the title of the article, as it is not necessarily necessary to use the investigation method and specify it in the title (Using Firm-Level Data Analysis). 

Materials and methods.   The authors described the investigation period of August and September 2020. The pandemic also took place in 2021 and until 2022. Why was the year 2021 at least not taken into account for the analysis? I suggest the authors to specify the software used for data analysis and arranging the figures.

Discussion. The authors make comparisons with studies from Africa, but do not specify that their study only concerns the implications on this continent. At least that is what emerges from their discussions in the text. I suggest the authors to make a clearer reference to these aspects.

Conclusions. The authors state their own contribution to the research study and make some future recommendations with implications for business in Africa. The authors also specify the limits of their research and outline some directions for future research.

Author Response

Overall responses

We strongly appreciate a thorough review and very rich inputs that contributed to improving the quality of the research. We have considered and addressed each comment and the response to each is presented below:

Abstract. The authors specify clearly what is the main objective of their research, the methodology used and the results obtained as a result of the research study.

Introduction. The authors present the general framework of the topic addressed and some of the implications of COVID-19 on SMEs from several countries. However, the authors do not structure this part of the Literature review on a certain geographical or continental area in order to make comparisons or to provide additional clarifications regarding the connection with the study undertaken by them. I suggest the authors to structure more clearly the implications of COVID-19 on SMEs on continental areas (for example Europe - Africa or America) to highlight certain results or implications of the epidemic with a greater emphasis on the wood business area, as the authors have pointed out in the main research objective.

Re: Thanks for the observation, while there was some information the pandemic effect on SMEs in Italy, Spain and China and Africa, mainly in Niageria and South Africa, I added three more literature or study findings on Africa (Line 160-165), East Asia (Line 462-4650, and Latin America (Line 481-485).

I also suggest the authors to underline this impact of COVID-19 on the wood business right in the title of the article, as it is not necessarily necessary to use the investigation method and specify it in the title (Using Firm-Level Data Analysis). 

Re: Thanks, while we intentionally omitted the wood subsector, to avoid narrowness to the title, I decided to keep the wood sector in the methodology section. This would allow visibility of the paper to the readers not only in agro-forestry. However, I’m fine to add a bit of nuance in the title and change it as: Analysis of Effects of the COVID-19 Pandemic on Small and Medium-Sized Enterprises (SMEs) in Rwanda Using Wood Firm-Level Data Analysis.

Materials and methods.   The authors described the investigation period of August and September 2020. The pandemic also took place in 2021 and until 2022. Why was the year 2021 at least not taken into account for the analysis? I suggest the authors to specify the software used for data analysis and arranging the figures.

Re: Thanks, it’s true that in 2021 covid-19 was still underway however, restriction measures were being lifted and economy was performing towards recovery. The survey  didn’t capture data up to 2021, thus in the analysis it’s a limitation as I mentioned towards the conclusion. I added information the software used for analysis. Line 216-17.

Discussion. The authors make comparisons with studies from Africa, but do not specify that their study only concerns the implications on this continent. At least that is what emerges from their discussions in the text. I suggest the authors to make a clearer reference to these aspects.

Re: Thanks for this comment, I left continental focus in the discussion because I didn’t necessarily have comprehensive information on the mechanics of Wood based SMEs in Africa. The narrative is mainly focusing on Rwanda but open to application elsewhere.

Conclusions. The authors state their own contribution to the research study and make some future recommendations with implications for business in Africa. The authors also specify the limits of their research and outline some directions for future research.

Re: Thanks very much

Reviewer 2 Report

Comments and Suggestions for Authors

The article deals with the economic and financial impact of the COVID-19 pandemic during the lockdown period in Rwanda. The literature background of the article can be improved with other relevant references. The research design was appropriate, and the methods were adequately described. I suggest to present the results of the research more clearly and complete the conclusions with the limitation of the research. I recommend revising the section on page 7: the content of Figure 1 and the text describing it do not contain the same values (e.g. 26% in the text, but 26.8 in the figure)

The overall merit of the article is average. The significance of the research can be describe in national, regional level for Rwanda.

 

Comments on the Quality of English Language

Minor editing of English language required.

Author Response

Comment 1: The article deals with the economic and financial impact of the COVID-19 pandemic during the lockdown period in Rwanda. The literature background of the article can be improved with other relevant references.

Response 1: Thanks for this highly relevant comment. While in the literature we had tried to be very selective because the COVID-19  pandemic effects were unique,  but we also tried to diversify the literature to cover both wood and non-wood-based SMEs. Also, our initial understanding was to put more efforts on the data analysis. To respond to the comments, we added few other relevant literatures to substantiate the case and give more information to the users. Specific additions are highlighted in Line 48-51, Line 59-65, Line 160-165, and Line 197 to 200.

Comment 2: The research design was appropriate, and the methods were adequately described. I suggest presenting the results of the research more clearly and complete the conclusions with the limitation of the research. I recommend revising the section on page 7: the content of Figure 1 and the text describing it do not contain the same values (e.g. 26% in the text, but 26.8 in the figure).

Response 2: Thanks, we updated the content on the figure 1 by including decimals, it was an omission and oversight, apologies!. We also added figure 4 to make the presentation on taxes more relevant to the audience and informative. We also made some modifications to the presentation of findings from regression. Also, in the conclusion of the article, we added  information on limitations of the methodology and the scope of the study see page 19

Comment 3: Minor editing of English language required.

Response 3: Proofreading of the paper was done and grammatical errors were fixed to ensure succinctness and greater comprehension across the documents.

Reviewer 3 Report

Comments and Suggestions for Authors

1.     The period “during lock-down” is fuzzy, considering the inertial influencing factors. Therefore, this period defined as such is not suitable. More precision is suggested.

2.     The strategies adopted by SMEs to reopen their business operations (Abstract) certainly refer to “post lock-down” period. This issue should be made clear across the paper.

 

3.     The author/s declare the “wood sector” (Keywords) of interest – because of “wood industry prevalence” (p.3, line 101). Then documented arguments should be provided.

4.  It is suggested to split the first section into two parts: (i) short introduction proper; and (ii) literature survey, focused on the studies relevant to the area and issues investigated (e.g., Sub-Saharan Africa, wood industry, SMEs, etc.), removing the less-relevant parts.

5.   It is recommended to restructure the methodology section in the logical order: research objective/s, method/s, instrument/s (and, of course, the target population under scrutiny – as the source of data).

6.    The algebraic signs in formula (1) should be revised.

7.   (2) is not a formula per se; and n=227 is the result of applying formula (1). Therefore, this paragraph should be reformulated and formulas re-numbered.

8.     Formula (4) should be explained.

9.     It is suggested to replace “sex” with “gender” across the paper (i.e. formula (6), line 227, etc.)

10.  Why the effect of employment (diff_employ) is marked ΔZ – as in formula (7) – and (diff_sale) is simply Y? The revision of formulas, symbols used is suggested.

11.  As the calculated sample size was 227, then why less SMEs (220) were investigated? This issue has to be addressed.

12.  In addition: Table 1 displays 217 respondents by size and 215 respondents by company age. This issue must be addressed, and possible misalignments should be explained.

13.  Were the target SMEs “interviewed” (line 248) or “surveyed” (as sample calculation suggested)? Uniform approach across the paper is strongly recommended.

14.  The abbreviations used (e.g., WPS, FRW) should be explained at their first use.

15.  The sample surveyed includes large companies, while the paper title and Abstract promise SMEs. Alignments in this respect are strongly recommended - across the paper.

16.  The sample structure by industry is not provided. It is strongly suggested to be.

17.  “220 SME operators” (line 248) were investigated is not quite correct – since among them there were large companies as well. Corrections are strongly recommended across the paper.

18.  In addition: as the term “operator” is too general and imprecise, it should be neatly declared what positions the respondents hold; and how qualified they were to be able to provide reliable answers.

19.  It is strongly suggested to split the last section into two parts: (i) policy recommendations addressed to the main stakeholders – as result of the research results; and (ii) conclusions - to declare to what extent the supposed research objective/s were reached.

 

20.  More attention to be paid to typos (e.g., et Al., – line 464; n,a – line 458, etc.) is suggested.

Comments on the Quality of English Language

Minor editing of English language is suggested.

Author Response

Paper’s title „Analysis of Effects of the COVID-19 Pandemic on Small and Medium-Sized Enterprises (SMEs) in Rwanda Using Firm-Level Data Analysis”:

- I think the title needs to be clarified, as the research sample only includes companies from the wood sector, I would personally add this information. The title could be, for example: "Analysis of the Effects of the COVID-19 Pandemic on SMEs in Rwanda from the Wood Industry / Processing Sector".

Re: Thanks, I added  information on wood, However I kept the information on data used for the readers to know the scope of the analysis.

- In the abstract, please change the tense to the present tense.

Re: Thanks, present tense was ensured where applicable. See highlighted text, however in some other context, we feel that it’s important to keep the past tense.

- In the introduction you write about the prevalence of the wood industry - lines 100-101. It would be worth mentioning some figures that show its major role in the economy.

Re: Thanks for the observation, I added a para with nationally available data on the contribution of agroforestry on GDP and on employment.

- Equation 1, in the square there should be a "=" sign.

Re: Thanks, we chose  to mean that sample is not a point estimation, rather even more values to minimum is acceptable, under the central limit theorem. However, we changed to avoid any confusion.

- In 2.1. you write about the simple random sample, which is actually derived from a selection of companies in the wood industry, not the whole country. Therefore, the results next to the estimation results can be valid if the selection was random and if the number of enterprises surveyed is above the minimum threshold. Note that they cover only one sector - so the results may only be valid for one sector.

Re: Yes, the sampling was random, and the total sample achieved is 244, however through modelling the sample varies, but both models and findings are robust. Additionally, the findings and their implications emphasize the wood based SMEs as a primary sector of interest, however, we believe that SMEs have several commonalities which imply that any issues facing enterprises in the wood sector also apply to others, thus allow generalizability. We further added in the limitation that future research directions need to focus on more large scale analysis which may include various SME types.

- Why don't you use natural logs for variables such as employment, turnover, etc. and (diff) difference in logs in the estimations? This would allow you to see the percentage changes in the dependent variables caused by 1 unit change in the independent variable. The estimation results would be clearer as the magnitude of the coefficients would be smaller. Secondly, the use of logs would allow some non-linear relationships to be captured.

Re: Thanks, this promoted me to try the log transformation, however given that the difference in sales, taxes and employment was depicting negative value, this necessitated to do more data transformation which like squaring the values, thus complicating interpretation of the findings. I also updated models’ description to include equations of the transformed indicators in log. This is a great insight as now easy to read coefficient. Much appreciation,  for easy reference, I also provided detailed Stata outputs for reference on coefficients. This will complement outlog2 tables as generated in stata.

- Have you tried other models? Because OLS is very sensitive to many data problems. One option would be to use a fixed effects regression to see if the results hold.

Thanks very much for this suggestion: I did not explore other models because, given the nature of the data—cross-sectional and capturing information on firms for few months on specific time intervals—I assumed that a fixed effects regression would not be appropriate. This also adds to the fact some variables were not captured with time aspect in mind, while others reflect changes in two intervals only, that why I referred to the difference (changes in period a before covid-19 and during the covid-19). However, to enhance the certainty of findings accuracy, I included a table with the results of normality test. Additionally, I performed a VCE bootstrap analysis to compare with robust standard errors. The results showed that the direction and significance of the models remain unchanged.

- Independent variables, do you have information on: foreign ownership? As it could show different responses to the pandemic crisis, e.g., https://doi.org/10.24136/eq.2023.024 (Thanks for the paper, I read it extensively and inspired my research interest in the subsequent papers

Re: Thanks, while this a very good observation and I appreciate this paper, however, we didn’t capture information on ownership by nationality in the data. Potentially this can be another variable of inquiry in the next analysis.  

- What about multicollinearity between variables such as: sales20 and employment, sales20 and tax_after, sales20 and . Surely you need to add information on the VIF or give information on the correlation coefficients in the appendix. This can be mainly attributed to log transformation which normalized the data and the use of heteroskedasticity-robust standard errors (VCE- Robust).

Re: Thanks, this is a great observation, I added a correlation matrix in the annex, and the test Variance Inflation factor shows generally a modest multicollinearity between the independent variables.

- Please describe the research sample (sectoral composition) in more detail and give some characteristics.

Re: Thanks, I added a bit of information on subsector composition.

Reviewer 4 Report

Comments and Suggestions for Authors

Paper’s title „Analysis of Effects of the COVID-19 Pandemic on Small ands Medium-Sized Enterprises (SMEs) in Rwanda Using Firm-Level Data Analysis”:

 

- I think the title needs to be clarified, as the research sample only includes companies from the wood sector, I would personally add this information. The title could be, for example: "Analysis of the Effects of the COVID-19 Pandemic on SMEs in Rwanda from the Wood Industry / Processing Sector".

- In the abstract, please change the tense to the present tense.

- In the introduction you write about the prevalence of the wood industry - lines 100-101. It would be worth mentioning some figures that show its major role in the economy.

- Equation 1, in the square there should be a "=" sign.

- In 2.1. you write about the simple random sample, which is actually derived from a selection of companies in the wood industry, not the whole country. Therefore, the results next to the estimation results can be valid if the selection was random and if the number of enterprises surveyed is above the minimum threshold. Note that they cover only one sector - so the results may only be valid for one sector.

- Why don't you use natural logs for variables such as employment, turnover, etc. and (diff) difference in logs in the estimations? This would allow you to see the percentage changes in the dependent variables caused by 1 unit change in the independent variable. The estimation results would be clearer as the magnitude of the coefficients would be smaller. Secondly, the use of logs would allow some non-linear relationships to be captured.

- Have you tried other models? Because OLS is very sensitive to many data problems. One option would be to use a fixed effects regression to see if the results hold.

- Independent variables, do you have information on: foreign ownership? As it could show different responses to the pandemic crisis, e.g. https://doi.org/10.24136/eq.2023.024

- What about multicollinearity between variables such as: sales20 and employment, sales20 and tax_after, sales20 and . Surely you need to add information on the VIF or give information on the correlation coefficients in the appendix.

- Please describe the research sample (sectoral composition) in more detail and give some characteristics.

Author Response

1. The period “during lock-down” is fuzzy, considering the inertial influencing factors. Therefore, this period defined as such is not suitable. More precision is suggested.

Re: Thanks, I added the actual meaning of lock-down to make it clear

2. The strategies adopted by SMEs to reopen their business operations (Abstract) certainly refer to “post lock-down” period.This issue should be made clear across the paper.

 Re: This is done and appreciated.

3. The author/s declare the “wood sector” (Keywords) of interest – because of “wood industry prevalence” (p.3, line 101). Then documented arguments should be provided.

Re: Thanks,  I added few more lines on wood sectors and their economic contribution to the economy.

4. It is suggested to split the first section into two parts: (i) short introduction proper; and (ii) literature survey, focused on the studies relevant to the area and issues investigated (e.g., Sub-Saharan Africa, wood industry, SMEs, etc.), removing the less-relevant parts.

Re: This is well appreciated; however, we removed the subtitle on the literature review to reduce the many titles in short sections. With the current structure, starting with the line 126 to 207, we refer to the review of essential literature. While we believe that all literature included is very selective and essential, we are also cognisant that this work is more on exploratory data analysis using the available data and we limited our investigations or meta-analysis of mainstream SMEs performance in the period of crisis, which can be another study in the future. I have added that in the next areas of academic inquiry.

5. It is recommended to restructure the methodology section in the logical order: research objective/s, method/s, instrument/s (and, of course, the target population under scrutiny – as the source of data).

Re: Thanks, the methods order was reviewed and updated appropriately

6. The algebraic signs in formula (1) should be revised.

Re: the formula 1 was updated accordingly.

7. (2) is not a formula per se; and n=227 is the result of applying formula (1). Therefore, this paragraph should be reformulated and formulas re-numbered.

Re: This was updated accordingly

8. Formula (4) should be explained.

Re: Additional explanations were added for better clarity in the model.

9. It is suggested to replace “sex” with “gender” across the paper (i.e. formula (6), line 227, etc.)

Re: Thanks, this was done across the tables of analysis

10. Why the effect of employment (diff_employ) is marked ΔZ – as in formula (7) – and (diff_sale) is simply Y? The revision of formulas, symbols used is suggested.

Re: Thanks for observations and advise: However, the letter choice was adopted as a matter of simplification to avoid confusion of letters if the same letter is used to each model. In this regard, Z value was used in place of Y of the previous formula, the same  was applied in Tax effect measurement, T to refer to the taxes effect.

11. As the calculated sample size was 227, then why less SMEs (220) were investigated? This issue has to be addressed.

Re: Thanks for highlighting this, in total we have 244 SMEs in the sample, I have updated. The error emanated in the analysis which didn’t effectively treat the missing values of some indicators. This was handled properly in the revised version.

12. In addition: Table 1 displays 217 respondents by size and 215 respondents by company age. This issue must be addressed, and possible misalignments should be explained.

Re: This is linked with the previous question; it’s now worked on.

13. Were the target SMEs “interviewed” (line 248) or “surveyed” (as sample calculation suggested)? Uniform approach across the paper is strongly recommended.

Re: Same as the above, number harmonization is correctly adjusted.

14. The abbreviations used (e.g., WPS, FRW) should be explained at their first use.

Re: Thanks: WPS explained and FRW explained under the table 1 Line 308

14. The sample surveyed includes large companies, while the paper title and Abstract promise SMEs. Alignments in this respect are strongly recommended - across the paper.

Re: Thanks, this is a great observation. It’s true that there are a few businesses which were surveyed and classified as large, however their influence on the overall paper is very minimal and we emphasized on on small and medium size enterprises. Having these large number also helped to serve as reference group across the analysis, especially when it comes to categorization.

15. The sample structure by industry is not provided. It is strongly suggested to be.

Re: Thanks, these questions, there are several descriptive tables which were omitted to keep the paper in the manageable size. However, I have added category/industry of wood SME as advised.

16. “220 SME operators” (line 248) were investigated is not quite correct – since among them there were large companies as well. Corrections are strongly recommended across the paper.

Re: Thanks, this was an error, proper presentation of number is ensured across the paper.

17. In addition: as the term “operator” is too general and imprecise, it should be neatly declared what positions the respondents hold; and how qualified they were to be able to provide reliable answers.

Re: This was addressed by changing from business operators to the business owners or their respective representatives.

18. It is strongly suggested to split the last section into two parts: (i) policy recommendations addressed to the main stakeholders – as result of the research results; and (ii) conclusions - to declare to what extent the supposed research objective/s were reached.

Re: Thanks, this is done the extent the research objective was achieved is added in the conclusion section.

20. More attention to be paid to typos (e.g., et Al., – line 464; n,a – line 458, etc.) is suggested.

Re: Thanks, the typos addressed as advised and another round of review to detect any alike issue as done.

Minor English editing

Re: Some English improvements were made where applicable.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thanks to the authors for making all the suggestions indicated.

Author Response

Comment 1:

Abstract. The authors specify clearly what is the main objective of their research, the methodology used and the results obtained as a result of the research study.

Introduction. The authors present the general framework of the topic addressed and some of the implications of COVID-19 on SMEs from several countries. However, the authors do not structure this part of the Literature review on a certain geographical or continental area in order to make comparisons or to provide additional clarifications regarding the connection with the study undertaken by them. I suggest the authors to structure more clearly the implications of COVID-19 on SMEs on continental areas (for example Europe - Africa or America) to highlight certain results or implications of the epidemic with a greater emphasis on the wood business area, as the authors have pointed out in the main research objective.

Response 1: Thanks for pointing out this, we agree with the reviewer, this comment has been addressed by adding literature on the implications of COVID-19 on SMEs. In the revised manuscript the changes are indicated in yellow. But specifically, we added the following in the revised paper: L42-45, page 1, page 2, line 64-68. and page 4 lines 184-189.   

Comment 2: I also suggest the authors to underline this impact of COVID-19 on the wood business right in the title of the article, as it is not necessarily necessary to use the investigation method and specify it in the title (Using Firm-Level Data Analysis). 

Response 2: Thank you for the comment. The title has been revised as “Analysis of Effects of the COVID-19 Pandemic on Small and Medium-Sized Enterprises (SMEs) in Rwanda Using Wood Firm level data”. See title page.

Comment 3: Materials and methods. The authors described the investigation period of August and September 2020. The pandemic also took place in 2021 and until 2022. Why was the year 2021 at least not taken into account for the analysis? I suggest the authors to specify the software used for data analysis and arranging the figures.

Response 3: Thank you for the comment. The survey was done in August and September 2020 and this was specified in the revised manuscript. Additionally, more than half of the all the cases recorded during the pandemic where recorded in 2020 (ref Cabore et al 2022) Stata software version 18 has been used in analysis and it has now been indicated in the revised manuscript. A reference was made in page 6, line 256- 258

Comment 4: Discussion. The authors make comparisons with studies from Africa, but do not specify that their study only concerns the implications on this continent. At least that is what emerges from their discussions in the text. I suggest the authors to make a clearer reference to these aspects.

Response 4: This comment has been addressed by adding references on previous studies from Africa as highlighted in the revised manuscript Line 63 and L68, Line 184-189, and Line 551-559, Line 582 -586.

Reviewer 3 Report

Comments and Suggestions for Authors

1.     The research method is not clearly mentioned – as it should be.

2.     The methodology section was not seriously revised as recommended.

3.     The confusion (to identification) of two different research methods (survey and interview) stands still.

4.     In addition, confusion between optimal sample size and minimum sample size stands still.

5.     The vagueness related to the sample (sample selection process and the criteria used for selection, etc.) stands still.

6.     The inclusion of large companies in the sample is in contradiction with the paper title.

7.     The reason provided among author/s’ responses (comparative analysis of SMEs against large companies) is not sustained by content, not even at basic, schematic level.

8.     The newly introduced dichotomy wood industry / forestry is neither documented nor analyzed. It should be.

9.     The promised post-covid strategies are neither identified nor analyzed. They should neatly be.

 

10.  The newly graphical materials added as appendices (diagrams called figures or annexes; numbered and un-numbered) are not referred to. The reason they were added is not transparent, and their value added is uncertain.

Comments on the Quality of English Language

Fair English

Author Response

Comment 1: Paper’s title Analysis of Effects of the COVID-19 Pandemic on Small and Medium-Sized Enterprises (SMEs) in Rwanda Using Firm-Level Data Analysis”:- I think the title needs to be clarified, as the research sample only includes companies from the wood sector, I would personally add this information. The title could be, for example: "Analysis of the Effects of the COVID-19 pandemic on SMEs in Rwanda from the Wood Industry / Processing Sector".

Response 1: Thanks, This comment was addressed by adding information on wood, However we kept the information on data used for the readers to inform the leader about the scope of the analysis. Now the title reads: “Analysis of Effects of the COVID-19 Pandemic on Small and Medium-Sized Enterprises (SMEs) in Rwanda Using Wood Firm Level Data”

Comment 2: In the abstract, please change the tense to the present tense.

Response 2: Thanks, present tense was ensured where applicable and the changes are highlighted in yellow. However, in some other context, we kept the past tense to keep the meaning of information. See for example lines 7, 13,

Comment 3: In the introduction you write about the prevalence of the wood industry - lines 100-101. It would be worth mentioning some figures that show its major role in the economy.

Response 3: Thanks for the observation, we added a paragraph with nationally available data on the contribution of agroforestry on GDP and on employment in Rwanda. See line page line 118-123.

Comment 4: Equation 1, in the square there should be a "=" sign.

Response 4: Thanks, Initially we chose  to mean that sample is not a point estimation, rather  presenting that more values to minimum is acceptable, under the central limit theorem. However, we changed to “=” sign.

Comment 5: In 2.1. you write about the simple random sample, which is actually derived from a selection of companies in the wood industry, not the whole country. Therefore, the results next to the estimation results can be valid if the selection was random and if the number of enterprises surveyed is above the minimum threshold. Note that they cover only one sector - so the results may only be valid for one sector.

Response 5: Thanks for the comments, this is to confirm that the sampling was random, and this was further clarified in the sampling methodology, and the total sample achieved is 244, however through modelling the sample varies, but both models and findings are robust. Additionally, the findings and their implications emphasize the wood based SMEs as a primary sector of interest, however, we believe that SMEs have several commonalities. This implies that any issues facing enterprises in the wood sector also apply to other businesses, thus allow generalizability. We further added in the limitation that future research directions need to focus more on large scale data analysis which may include various SME types, see page 19, line 634-642.

Comment 6: Why don't you use natural logs for variables such as employment, turnover, etc. and (diff) difference in logs in the estimations? This would allow you to see the percentage changes in the dependent variables caused by 1 unit change in the independent variable. The estimation results would be clearer as the magnitude of the coefficients would be smaller. Secondly, the use of logs would allow some non-linear relationships to be captured.

Response 6: Thanks, this promoted us to try the log transformation, however given that the difference in sales, taxes and employment was depicting negative value, this necessitated to do more data transformation mainly squaring the values. We also updated models’ description to include equations of the transformed indicators in log. This is a great insight as now easy to read coefficient are displayed in the model. Additionally,  we provided detailed Stata outputs for reference on coefficients in the annex of the paper. This will complement outlog2 tables presented in the main paper, changes were applied formula 7, 9 and 11, the same changes were observed in figure 6 page 15.

Comment 7: Have you tried other models? Because OLS is very sensitive to many data problems. One option would be to use a fixed effects regression to see if the results hold.

Response 7: Thanks very much for this suggestion: we intentionally did not explore other models given the nature of the data—cross-sectional and capturing information on firms for few months on specific time intervals— we assumed that a fixed effects regression would not be appropriate. This also adds to the fact some variables were not captured with time variance aspect in mind, while others reflect changes in two intervals only, that why we referred to the difference (changes in period a before covid-19 and during the covid-19). However, to enhance the certainty of findings accuracy, we included a table with the results of normality test. Additionally, we performed a VCE bootstrap analysis to compare with robust standard errors. The results showed that the direction and significance of the models remain unchanged. See the annex statistical tables on pages 23-25

Comment 8: Independent variables, do you have information on foreign ownership? It could show different responses to the pandemic crisis, e.g., https://doi.org/10.24136/eq.2023.024

Response 8: Thanks, while this a very good observation and we appreciate this paper, however, we didn’t capture information on ownership by nationality in the data. Potentially this can be another variable of inquiry in the next analysis.

Comment 9: What about multicollinearity between variables such as: sales20 and employment, sales20 and tax_after, sales20. Surely you need to add information on the VIF or give information on the correlation coefficients in the appendix. This can be mainly attributed to log transformation which normalized the data and the use of  heteroskedasticity-robust standard errors (VCE- Robust).

Response 9: Thanks, this is a great observation, we added correlation coefficient in the annex and the test Variance Inflation factor (VIF), shows a modest multicollinearity between the independent variables, and the results are robust. See annex under page 26.

Comment 10: Please describe the research sample (sectoral composition) in more detail and give some characteristics.

Response 10: Thanks, we added more information SMEs characteristics including subsector composition to lay the foundation of understanding SMEs characteristics. See page 13 line 228-234.

Reviewer 4 Report

Comments and Suggestions for Authors

The paper has been greatly improved and enhanced. The paper is well structured. The information is presented in a clear way. The discourse is interesting. However, I would like to comment on one thing that I think the authors have left out, namely foreign capital. Perhaps in the discussion section or in the concluding section, the potential role of FDI in building resilience, together with important sectoral effects, could be sketched out a bit, e.g. https://doi.org/10.24136/eq.2023.024

Small things:

-            Please correct Figure 5. Employment … (there is a typo on the chart- employment before)

Author Response

Comment 1: The period “during lock-down” is fuzzy, considering the inertial influencing factors. Therefore, this period defined as such is not suitable. More precision is suggested.

Response 1: Thanks, we added the actual meaning of lock-down to make it clear, see page 1 line 6

Comment 2: The strategies adopted by SMEs to reopen their business operations (Abstract) certainly refer to “post lock-down” period. This issue should be made clear across the paper.

Response 2: This has been updated by clearly articulating the re-opening strategy in the main texts including literature review and the conclusion of the paper. See page 18, line 578-583 and line588-592

Comment 3: The author/s declare the “wood sector” (Keywords) of interest – because of “wood industry prevalence” (p.3, line 101). Then documented arguments should be provided.

Response 3: Thanks, we added few more sentences on wood sectors and their economic contribution to the economy. See page 3 line

Comment 4: It is suggested to split the first section into two parts: (i) short introduction; and (ii) literature survey, focused on the studies relevant to the area and issues investigated (e.g., Sub-Saharan Africa, wood industry, SMEs, etc.), removing the less-relevant parts.

Response 4: This is well appreciated; however, we initially removed the subtitle on the literature review to reduce the many titles in short sections. With the current structure, starting with the line 126 to 207, we refer to the review of essential literature. While we believe that all literature included is very selective and essential, we are also cognisant that this work is more on exploratory data analysis using the available data and we limited our investigations or meta-analysis of mainstream SMEs performance in the period of crisis, which can be another study in the future. We have added that in the next areas of academic inquiry. Page  19, line 635,

Comment 5: It is recommended to restructure the methodology section in the logical order: research objective/s, method/s, instrument/s (and, of course, the target population under scrutiny – as the source of data).

Response 5: Thanks, the methods order was reviewed and updated appropriately and as advised, page 5, line 213 and added lines 242-244.

Comment 6: The algebraic signs in formula (1) should be revised.

Response 6: The formula 1 was revised accordingly.

Comment 7: (2) is not a formula per se; and n=227 is the result of applying formula (1). Therefore, this paragraph should be reformulated and formulas re-numbered.

Response 7: Thanks very much this this was updated accordingly, now the figures are harmonized including numbering.

Comment 8: Formula (4) should be explained.

Response 8: Additional explanations were added for better clarity of the model. See line 273, and due to additional specifications, the formula is not 6.

Comment 9: It is suggested to replace “sex” with “gender” across the paper (i.e. formula (6), line 227, etc.)

Response 9: Thanks, this was done across the paper.

Comment 10: Why the effect of employment (diff_employ) is marked ΔZ – as in formula (7) – and (diff_sale) is simply Y? The revision of formulas, symbols used is suggested.

Response 10: Thanks for observations and advise: However, the letter choice was adopted as a matter of simplification to avoid confusion of letters if the same letter is used to each model. In this regard, Z value was used in place of Y of the previous formula, the same  was applied in Tax effect measurement, T to refer to the taxes effect. See page 7 formula 11 and

Comment 11:  As the calculated sample size was 227, then why less SMEs (220) were investigated? This issue has to be addressed.

Response 11: Thanks for highlighting this, in total we have 244 SMEs in the sample, we have updated. The error emanated in the analysis which didn’t effectively treat the missing values of some indicators. This was handled properly in the revised version. See page 5 line 229-234

Comment 12: In addition: Table 1 displays 217 respondents by size and 215 respondents by company age. This issue must be addressed, and possible misalignments should be explained.

Response 12: This is linked with the previous question; it’s now worked on by correctly updating the sample size. Line 229-234

Comment 13: Were the target SMEs “interviewed” (line 248) or “surveyed” (as sample calculation suggested)? Uniform approach across the paper is strongly recommended.

Response 13: Same as the 12, number  is adjusted, we have used surveyed across the board.

Comment 14:  The abbreviations used (e.g., WPS, FRW) should be explained at their first use.

Response 14: Thanks, WPS explained and FRW explained under the table 1 Line 308.

Comment 15: The sample surveyed includes large companies, while the paper title and Abstract promise SMEs. Alignments in this respect are strongly recommended - across the paper.

Response 15: Thanks, this is a great observation. It’s true that there are a few businesses which were surveyed and classified as large, however their influence on the overall paper is very minimal, and we emphasized on small and medium size enterprises. Having larger companies also helped to serve as reference group across the analysis, especially when it comes to the categorization. We have however, removed making a reference to  large companies’ findings in the main report to be consistent.

Comment 16: The sample structure by industry is not provided. It is strongly suggested to be.

Response 16: Thanks, there are several descriptive tables which were omitted to keep the paper in the manageable size. However, we addressed these comments by adding category/industry of wood SME as advised and the distribution of the surveyed SMEs by districts.

Comment 17: “220 SME operators” (line 248) were investigated is not quite correct – since among them there were large companies as well. Corrections are strongly recommended across the paper.

Response 17: Thanks, this was an error, a proper presentation of number is ensured across the paper.

Comment 18: In addition: as the term “operator” is too general and imprecise, it should be neatly declared what positions the respondents hold; and how qualified they were to be able to provide reliable answers.

Response 18: This was addressed by changing from business operators to the business owners or their respective representatives. See line 220

Comment 19: It is strongly suggested to split the last section into two parts: (i) policy recommendations addressed to the main stakeholders – as result of the research results; and (ii) conclusions - to declare to what extent the supposed research objective/s were reached.

Response 19: Thanks, this is done and the extent at which the research objective was achieved is added in the conclusion section. See line 602

Comment 20: More attention to be paid to typos (e.g., et Al., – line 464; n,a – line 458, etc.) is suggested.

Response 20: Thanks, the typos were addressed as advised and another round of review to detect any alike issue as done.

Comment 21: Minor English editing

Response 21: Some English improvements were made where applicable throughout the report.

Round 3

Reviewer 3 Report

Comments and Suggestions for Authors

All suggestions made already (which remained answerless).