Analysis of the Determining Factors for the Renovation of the Walloon Residential Building Stock
Round 1
Reviewer 1 Report
the main sources from which the data are retrieved has different coverage rates: EPC are available for 495.470 dwellings, while the breakdown by typology is on the entire building stock (1.615.774 dwellings), as the household breakdown by income
So, the retrieved EPCs cover 25% only of total dwelling stock. Why EPC are not availabe for the remaining 3/4 of the stock? Does this little share representative of the entire stock? Please provide some comments about this within the paper
Author Response
Please see the attachment
Author Response File: 
Author Response.docx
Reviewer 2 Report
The presented paper characterizes the Walloon residential building stock and analyzes the existing correlations between the stock's technical data and
its occupants socio-economic data. The authors highlight the importance of focusing on renovation strategies for particular types of buildings. The paper is well structured, clearly written, the analysis is based on actual data statistical analysis of the Walloon region.
In the paper conclussion section is missing.
As the paper is not very original as well as conlussions confirm the correclations which are already known from other studies, authors must compare the outcomes with the results of the other studies and clearly state where is the novelty of their study.
Overall, the paper is interesting, well written, but authors must emphasize the novelty.
Author Response
Please see the attachment
Author Response File: Author Response.docx
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
Decision:
In my view, this paper lacks the characteristics generally expected from a journal publication. The main reason is that the paper does not contribute to the existing body of knowledge. For example, it is quite clear that apartments are generally more energy-efficient than single-family houses when the authors do not control for the size, age, and quality of buildings. Also, it is quite clear that low-income households tend to live in low-quality houses that tend to be less energy-efficient than houses occupied by more affluent households. In addition, the paper does not provide specific or critical policy solutions, other than suggesting that a targeted renovation policy needs to be established with a particular focus on low-income households, which is, again, an obvious fact. I suggest that the authors rethink and reshape the literature review to determine what is known and what is unknown.
Major comments:
The content is generally fragmented, lacks focus, and is very hard to follow. The paper should be entirely rewritten, and its language, grammar, and punctuation should be revised, corrected, and improved, considering:
- Problem Statement: What is the research problem? What is unknown in the existing body of knowledge that the authors want to make known/clarify by performing this analysis?
- Rationale/Objectives: What are the objectives of this work? How successful is this research in achieving its objectives?
- Contributions: What are the three (or more) contributions that this work has for other researchers in the field of energy efficiency?
- Analysis: significant revisions are required. Before describing the regression analysis, the authors should provide a table the summarizes the statistics of the dependent and independent variables, including minimum, maximum, mean, and a description of the way the variables are operationalized. The regression analysis itself should be presented in a table that includes all the variables, coefficients, standard errors, and p-values. The unit of analysis and sample size is not clear. When was the data collected? The authors mention that the sample size is 495,470 EPC for 1,615,774 registered dwellings. What does this mean? If one of these numbers is the sample size, this means that the unit of analysis is building. If this is the case, why did the authors use the share of income, housing type, etc. at the province level? More importantly, if the unit of analysis is building, why did not the authors account for other important building characteristics like the age of construction, house size, and construction quality?
Minor comments:
Line 50: What does SBR stand for? Please define all the acronyms before using them.
Line 60: What was the previous role of housing renovation policies?
Line 80: Why is multi-collinearity important here? Multi-collinearity can be easily avoided in the analysis.
Line 185: Sample size has a formula that should be applied ‘in this analysis’ to confirm its representativeness.
Author Response
In my view, this paper lacks the characteristics generally expected from a journal publication. The main reason is that the paper does not contribute to the existing body of knowledge. For example, it is quite clear that apartments are generally more energy-efficient than single-family houses when the authors do not control for the size, age, and quality of buildings. Also, it is quite clear that low-income households tend to live in low-quality houses that tend to be less energy-efficient than houses occupied by more affluent households. In addition, the paper does not provide specific or critical policy solutions, other than suggesting that a targeted renovation policy needs to be established with a particular focus on low-income households, which is, again, an obvious fact. I suggest that the authors rethink and reshape the literature review to determine what is known and what is unknown.
Thank you very much, Reviewer 1. We appreciate your time and effort, and we are very happy to receive such constructive feedback. We did our best to address all your comments. Please follow our modifications explained point by point. All changes are identified through tracked changes.
Major comments:
The content is generally fragmented, lacks focus, and is very hard to follow. The paper should be entirely rewritten, and its language, grammar, and punctuation should be revised, corrected, and improved, considering:
We have thoroughly revised the document and verified the English with professional proofreading. Attached you can find the copy of proofreading certificate.
- Problem Statement: What is the research problem? What is unknown in the existing body of knowledge that the authors want to make known/clarify by performing this analysis?
We agree. We reformulated the sentence to make sense. See Line 119.
- Rationale/Objectives: What are the objectives of this work? How successful is this research in achieving its objectives?
We agree. We reformulated the sentence to make sense. See Line 123 and 463.
- Contributions: What are the three (or more) contributions that this work has for other researchers in the field of energy efficiency?
We agree. We reformulated the sentence to make sense. See Line 492.
- Analysis: significant revisions are required. Before describing the regression analysis, the authors should provide a table the summarizes the statistics of the dependent and independent variables, including minimum, maximum, mean, and a description of the way the variables are operationalized.
We agree. See table 2.
The regression analysis itself should be presented in a table that includes all the variables, coefficients, standard errors, and p-values.
We agree. See table 3.
The unit of analysis and sample size is not clear. When was the data collected?
The year of each data collection is given in Table 1. It depends on their origin.
The authors mention that the sample size is 495,470 EPC for 1,615,774 registered dwellings. What does this mean? If one of these numbers is the sample size, this means that the unit of analysis is building.
We agree. We precise the methodology. See Line 190.
If this is the case, why did the authors use the share of income, housing type, etc. at the province level?
We agree. We precise the methodology. See Line 208.
More importantly, if the unit of analysis is building, why did not the authors account for other important building characteristics like the age of construction, house size, and construction quality?
We agree. We complete the discussion to make sense. See Line 505.
Minor comments:
Line 50: What does SBR stand for? Please define all the acronyms before using them.
We agree. We correct the error the formulation. See Line 51.
Line 60: What was the previous role of housing renovation policies?
We agree. We reformulated the sentence to make sense. See Line 51.
Line 80: Why is multi-collinearity important here? Multi-collinearity can be easily avoided in the analysis.
Maybe the comment is not understood, but the multicollinearity remains limited (see the calculation of the VIF line 405) despite the analysis of characteristics that could be subject to it.
Line 185: Sample size has a formula that should be applied ‘in this analysis’ to confirm its representativeness.
We preferred to rely on a study comparing the representativeness of different characteristics of the EPC database with the known characteristics of the entire built stock. Indeed, we are well aware that the mode of data collection is not at all random. It could have very clearly favored this or that building (despite the large number of buildings far exceeding the calculation of the required sample size)*.
*n = t² × p × (1-p) / m² = 2.58² × 0,5 × 0,5 / 0,01² = 16.641 <<< 495.470 for the worst case
Thank you very much Reviewer 1.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Overall, well written and interesting. I like your approach of using official statistics as a basis for this type of analysis. I do, however have some comments.
First, simple formalities:
Line 162: reference missing
229: you’re writing in future tense: “The results will be tested for hypothesis validity”
In figure 1, “recommendation” is misspelled.
Fig 3, 4 and 6: there are some percentage figures in the diagrams, but they seem completely random. Or?
Then, I would like to see some clarifications:
Line 340-41: Four-sided houses in small towns are mostly owner-occupied, and more than the average number are rented…Does this make sense? I suppose owner-occupied + rented = 100 %? How can then the number of rented houses be higher than average, when "most" of them are owner-occupied? Because the number of owner-occupied are lower than average? Please clarify
I’m confused about the geometries. In line 217-18 it says: “the first three explanatory variables are the proportion of apartments, semi-detached houses, and detached houses”.
In Line 223, it says: “the proportions of detached houses are redundant”
In Figure 4, there are four categories: apartments, detached, semi-detached houses, and terraced houses. In figure 7, there are again three categories: apartments, semi-detached houses, and detached houses. Is it the proportions of terraced houses that are redundant?
In figure 1, showing the “Study conceptual framework”, on the level of “Instrumental development”, there is a box labeled “Simulation of the energy efficiency, according to other factors”. On line 364, you talk about an estimated 721,83 kWh/m2y. I cannot trace where this comes from, and this is the first time energy efficiency is mentioned. In section 3.4.2. (line 3.4.2 and onwards) you talk about "calculating the total impact", without specifying what that is. In table 3, there are figures, but no unit. I assume, based on the last sentence, this is related to energy efficiency? These figures are in the range of 330 up to 458. How does this relate to the figure 721,83 kWh/m2y?
I find your work interesting, and your conclusions very relevant, but I miss vital information on the simulation part.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The paper focuses on a current topic with relevant social implications and potential effects on public policies
The proposed method correctly takes into account the difficulty of correlating information of a different nature, and the limitations due to privacy concerns
The topic is framed by a comprehensive and well-structured literature review
This reviewer just suggest the following points of attention to enhance the paper quality:
Line 22: the correct symbol for kilowatt-hour in SI is: kWh. So: kWh/m2y, instead of kWH/m2y
Lines 42 to 45: To make the comparison more immediate, it would be appropriate to indicate the position of Belgium with respect to the European averages (share of direct energy consumption of the residential sector, age of the building stock and index of o energy improvements penetration)
Line 113: <<…the social aspects, which are essential in Belgium>>. Why in Belgium more that somewhere else? Provide a short explanation of this
Lines 145-146: <<There is a correlation between the housing geometry and social aspects, such as household income or property ownership [41]>>
Briefly summarize the outcomes of the cited research [reference 41], in order to allow understanding how this correlation operates. Does a direct link between building geometry and user income exists? Or a third element must be considered too, concerning the building location? Furthermore: is it a presupposition (in this case, it must be clarified better) or a research hypothesis, which will be treated later in the paper?
Lines 147 to 149: << The energy regulation for buildings is different (in Walloon region) from other regions in Belgium (Brussels and Flanders) and other countries because the application of the European EPBD was specifically translated in the Walloon regulation as the Performance 149 Energétique des Bâtiments (=EPC)>> Explain what the difference is
Line 162: Reference source not found
Lines 165 to 170: what is the coverage rate of EPC issued with respect to the total of the building stock? Specify
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 4 Report
The paper is very well structured, all the parts of the paper are very well written and raise no additonal questions. In general paper is very interesting and despite that it has average scientific novelty, as no new methods are proposed, performed analysis is usefull and can be applied not just for Walloon building stock, but also has potential to be applied in other regions.
I have just one proposition for authors to reconsider the Title, as content is more significant compared to the Title. Minimum what can be change - instead of "characterization" "analysis" would sound more scientific and correspondance to the content would be better. But I also would like to propose to reconsider the title to make it more informative and attractive to the reader.
Some wording like "statistical analysis", "mixed-model", etc. could be used in the title.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
I do not think this paper is successful in presenting any novel results. Nor does it have clearly stated, significant research objectives and supporting rationale. The content of the discussion section is inadequate for a journal paper. There are still many writing errors and omissions in the text. I cannot accept this paper in this form. The following includes comments on the discussion section:
Line 476 “The objective of this research was to analyze and quantify the existing correlations between the technical data of the residential stock and the socio-economic data of the inhabitants of this stock, on the scale of the Walloon region.”
Comment: This research does not have a clearly stated research objective. Therefore, I cannot accept it in this form. A research objective can be analyzing a correlation between two or more specific variables based on a stated need or reason. For example, researchers may analyze the correlation between the density of housing stock and energy consumption to see if an increase in density can decrease energy consumption when all other variables are held constant. A research objective cannot be simply defined as quantifying correlations between two or more sources of data, without having a specific reason.
Line 483: “The correlation part shows that apartments are correlated with improved energy efficiency, and semi-detached houses are correlated with lower energy efficiency.”
Comment: This is not a new finding. Rather, this is a fact most of us already know. Apartments are often small and dense, having high volume to exposed surface ratio compared to detached and semi-detached single-family houses. Therefore, we do not expect apartments to be less energy efficient than single-family houses.
Line 485: “Most importantly, MLR provides more surprising results. Literature on split incentives suggested that a decrease in the ownership rate could be accompanied by a decrease in energy efficiency. On the contrary, MLR highlights the correlation between increased ownership rate and decreased energy efficiency.”
Comment: This is not surprising. Past research says there could be a positive correlation between the two. The reference [44] is missing in the list of references so it is not clear where this claim is coming from.
Line 489: “More specifically, the study shows that very low incomes (10,000–20,000€ of net taxable income) occupy much less efficient housing than the rest of the population. This correlation diminishes very quickly for higher-income groups.”
Comment: Again, this is not a new finding. No one expects low-income populations to live in high-quality, energy-efficient houses. It is already clear that renovation policies should prioritize low-income households because energy spending makes a significant percentage of low-socioeconomic populations’ income and most populations at the lowest income quintal are not able to renovate without receiving some sort of assistance.
