Have Extreme Events Awakened Us?

Round 1

Reviewer 1 Report

The paper titled “Have Extreme Events Awakened US? Concepts and World Evidence” presents a study based on their conceptual-empirical model to investigate whether extreme natural events affect the decision of the policymakers to make more stringent environmental regulations in different countries. The study is interesting and well presented. I have some comments and concerns listed below. The authors are required to address them adequately before any recommendation can be made.

1. In this study, the authors assumed that climate change is a reaction to the unsustainable economic activities of mankind. Climate change and extreme natural events can even be the consequences of non-anthropogenic attributes. The authors need to comment on how to take that into account in their models and present the results accordingly.
2. The basis of the empirical equation (1) is to be explained. What is the basis for adopting such a linear model? The calibration and validation studies of such a model are to be presented in detail with real-life data.
3. What is the meaning of negative values in Tables 1 and 2?
4. Why use US in capitals in the title?
5. Moreover, I believe the conclusions drawn based on a very simplistic empirical model presented here are not well substantiated. The results are also not that detailed to arrive at the conclusions drawn here. The authors need to present more rigorous results with a strongly validated model to arrive at their conclusions. In the present form, the study does not look convincing.

Author Response

1. In this study, the authors assumed that climate change is a reaction to the unsustainable economic activities of mankind. Climate change and extreme natural events can even be the consequences of non-anthropogenic attributes. The authors need to comment on how to take that into account in their models and present the results accordingly.

Response: It must be noted that the primary objective of this analysis is to examine if extreme events have motivated policy makers to make enforce more stringent environmental regulations. So, this analysis does not whether the extreme events were driven by anthropogenic or natural climatic changes. A clarification has been added to the text.  

1. The basis of the empirical equation (1) is to be explained. What is the basis for adopting such a linear model? The calibration and validation studies of such a model are to be presented in detail with real-life data.

Response: The basis of this model is meant to be explained in the conceptual model (developed by the authors), linking different socio-economic-environmental elements of this study to each other and discussing intuitions behind that. A paragraph has been added to the empirical section elaborating this point of view. However, the main argument of this study is not to measure the exact point estimate of those relationships but only to control for their impacts on the main analysis, which is the impact of extreme events on the environmental policies. Thus, the main focus of this abstract analysis is revealing such connections, not the linkage of others, and controlling for them to remove/reduce the omitted variables’ biases. 

1. What is the meaning of negative values in Tables 1 and 2?

Response: When variables volume increases, the dependent variable may decrease (such as the possible correlation). However, most of those values are insignificant with high standard errors meaning there is no such (meaningful) correlation. The tables’ formats have been revised to reflect the primary purpose of this analysis: the impact of extreme events on environmental policies.  

1. Why use US in capitals in the title?

Response: It was wrongly capitalized; it’s been changed to lower case (Us) to avoid confusion with the United States.   

1. Moreover, I believe the conclusions drawn based on a very simplistic empirical model presented here are not well substantiated. The results are also not that detailed to arrive at the conclusions drawn here. The authors need to present more rigorous results with a strongly validated model to arrive at their conclusions. In the present form, the study does not look convincing.

Response: This abstract analysis aims to establish a convincing question on the relationship between environmental policies’ origins/urgencies and the extreme events, not to claim the exact point estimate—therefore precise causation—of this relationship. And, of course, a more in-depth empirical analysis is needed to calculate such impact. That’s the initial reason we submitted this work as a short communication, not a comprehensive article. A clarification has been added to the conclusion to avoid such claims.

Reviewer 2 Report

The topic of this study is interesting. The paper  use  yearly panel data from 1990 to 2017 for more than 40 OECD countries to examine such a relationship. Although this paper has innovative topic, clear logic and reasonable conclusion, it still remains a number of loopholes. I suggest a major revision, some comments and suggestions are given below for the authors to improve the paper:

1.The literature review part is too weak, the author has not given a relatively comprehensive review on the existing literature, and therefore it is hard to highlight the contribution of this research comparing with existing research.

2.Figure 1 is not clear, you need to redo it.

3.Please check the whole paper, make sure the language fulfill the requirement of the journal.

4.Some fresh papers can be added as references:

Sun, H., Samuel, C.A,  Amissah, JCK, Taghizadeh-Hesary, F., Mensah, IA., 2020. Non-linear nexus between CO2 emissions and economic growth: A comparison of OECD and B&R countries,  Energy 212, 118637. https://doi.org/10.1016/j.energy.2020.118637.
Sun H., Edziah B K., Sun C., Kporsu A K. Institutional quality and its spatial spillover effects on energy efficiency，Socio-Economic Planning Sciences，2021，101023. https://doi.org/10.1016/j.seps.2021.101023.

 

Author Response

1. The literature review part is too weak, the author has not given a relatively comprehensive review on the existing literature, and therefore it is hard to highlight the contribution of this research comparing with existing research.

Response: The literature review is abstract since we aimed for a short communication study, not a comprehensive analysis/paper. To address the reviewer’s concern, more references and explanations have been added to the intro to improve this section.

1. Figure 1 is not clear; you need to redo it.

Response: This figure aims to illustrate potential connections between the used elements in this study. More explanations have been added to the conceptual section to clearer such connections.

1. Please check the whole paper, make sure the language fulfills the requirement of the journal.
2. Response: A comprehensive grammar check has been conducted to address such caveat; and text has been edited thoroughly. 
3. Some fresh papers can be added as references:

Sun, H., Samuel, C.A,  Amissah, JCK, Taghizadeh-Hesary, F., Mensah, IA., 2020. Non-linear nexus between CO2 emissions and economic growth: A comparison of OECD and B&R countries,  Energy 212, 118637. https://doi.org/10.1016/j.energy.2020.118637.
Sun H., Edziah B K., Sun C., Kporsu A K. Institutional quality and its spatial spillover effects on energy efficiency，Socio-Economic Planning Sciences，2021，101023. https://doi.org/10.1016/j.seps.2021.101023.

Response: These papers have been added to the introduction, strengthening our point of view.

Reviewer 3 Report

Review:

 Have Extreme Events Awakened US? Concepts and World Evidence 

The MS is vaguely written and has many flaws. It appears that they didn’t even pay attention to the basics:

1.     Grammartical error are seen in the paper. Editor needs to pay attention. 

2.     All of a sudden a new style font has been used. Line 66-

3.     Has OCED been defined before abbreviated? 

4.     Line 53_ six million deaths due to COVID- a reference or citation to a website is needed.

5.     Please provide a link to the database used.

6.     I don’t think that the model can be so simple. I dont think its so simple, what about covariability among the indep parameters? What about overfitting? Any regularizations not considered? Why not? Did they test the linearity - non linearity? Why don’t they use some activation functions to check that?

7.     Results: the authors can see something in the table as in line 158. I dont see it. please explain the table in a proper and clear way.  What dop these values and their signs mean. Explain every row and column. That’s not how to write a paper. Impossible to understand from that paragraph what the table is about. Same for table 2.

8.     Line 221: it’s not a paper. It’s a manuscript. 

Author Response

1. Grammartical error are seen in the paper. Editor needs to pay attention. 

Response: A grammar check has been conducted, and errors have been removed.

1. All of a sudden, a new style font has been used. Line 66-

Response: The font has been changed to the original format.

1. Has OCED been defined before abbreviated?

Response: The definition is added to the abstract. 

1. Line 53_ six million deaths due to COVID- a reference or citation to a website is needed.

Response: A citation from World Health Organization has been added.

1. Please provide a link to the database used.

Response: Before the reference section, two links are provided in the data availability statement.

1. I don’t think that the model can be so simple. I don’t think it’s so simple, what about covariability among the indep. parameters? What about overfitting? Any regularizations not considered? Why not? Did they test the linearity - non linearity? Why don’t they use some activation functions to check that?

Response: A robustness check has been added (Table 3), addressing concerns on co-variability between death rates and the number of incidents, non-linearity, test for overfitting, and GLM-panel estimation to validate the conclusion, partially. We have also performed/tested other forms to derive the current state, as bolded in the text.

It’s worth mentioning that since the death rates are the consequences of extreme events. Therefore, including number of the events and deaths simultaneously may not be desirable since death is the mechanism between extreme events and the dependent variable (environmental policies). One absorbs the other impact as a channel to the dependent variable.

On regularizations, the dependent variable is environmental policies which reflect environmental regulations. Also, we added political views to capture and control the policy-makers point of view within the process.

1. Results: the authors can see something in the table as in line 158. I don’t see it. please explain the table in a proper and clear way.  What dop these values and their signs mean. Explain every row and column. That’s not how to write a paper. Impossible to understand from that paragraph what the table is about. Same for table 2.

Response: A paragraph has been added explaining the percentage change calculation. The tables’ formats have also been changed to reflect the primary purpose of this analysis—the impact of extreme events on environmental policies—; the other elements are labeled control variables since we included them (in the empirical section) to decrease the biases due to omitted variables, not to estimate their effects on environmental policies.   

1. Line 221: it’s not a paper. It’s a manuscript. 

Response: It’s been revised.

Round 2

Reviewer 1 Report

The authors have addressed the queries adequately. The paper can now be accepted for publication.

Author Response

Thanks for your constructive comments.

Reviewer 2 Report

Now the revised version is better, so it can be accepted.

Author Response

Thanks for your constructive comments.

Reviewer 3 Report

Please do some typo checks, for example- line 200 has two ".." The paper is not good and the authors don’t show the interdependency of the parameters clearly. Its flawed to assume that the relationships between the features are independent. after reading the paper two times, its still not clear to me. I cant provide same comments again. But, yes- the relationship that they come up with is wrong/flawed because they didn’t check the interdependency and didn’t clearly explain the non linearity. On non linearity, they have provided some values, but didn’t clearly explain what that is.

the paper is short and can be improved on focusing on the interdependency of the features and how coefficient shrinkage method like L1 and L2 regularization can help in the equation.

 

if they have already done that, please state clearly. from the current state, its not clear at all.

 

 

Author Response

We found and fixed more than two dozen typos throughout the manuscripts. We did not claim that all the variables are independent of each other. On the contrary, we tried to show that all of them are connected directly and indirectly and can influence each other (figure one). In addition, the only independent variable in the econometrics model is extreme events, and the rest depend on each other. Attached is the DAG diagram of the empirical model.

Control variables are included to reduce the biases due to omitted variables. Still, their exact impact is not the purpose of this study and requires a separate and advanced investigation and may answer different questions than we are looking for in this Communication. Shrinkage methods are vastly used in the prediction models, whereas in this abstract analysis, we do not predict anything. Instead, we’re constructing a basic causal model, focusing on the exogeneity of independent variables, taking advantage of the fixed effects in the panel, including the control variables, and performing a couple of robustness checks to make sure what we’re claiming is more than a correlation, but close to causation (using well-known methods discussed in “Mostly harmless econometrics: An empiricist’s comparison” by Angrist & Pischke, 2009 among others like Morgan & Winship, Rosenbaum, etc.). Alternatively, on the referee’s note, we added a selection test on observed and unobserved variables (same idea as in the shrinkage method), proposed by Altonji et al. (2005), depicted in Table 4 in the text. Also, more clarifications have been added about the econometrics model, including the presented DAG diagram. We also showed in Table 4 that including the non-linear forms of the independent variable wouldn’t change the primary results.  In the end, we need to emphasize our sincere gratitude to the referee, who without, we wouldn’t have had such a polished study with an improved methodology.

Author Response File: Author Response.pdf