Impacts of Layoffs and Government Assistance on Mental Health during COVID-19: An Evidence-Based Study of the United States
Round 1
Reviewer 1 Report
Lines 15- 17 and 167- With the sentence, "Our statistical 15 analysis uses the ordinal regression model (ordered logit model) that takes both the week and 16 regional factors into consideration to control for potential time effects", what effect would the week have on mental health? This is referenced also in line 167, but could be expanded upon.
Lines 43-44 - Good discussion. It may be worth considering that since there are racial and ethnic disparities in COVID-19 sickness, hospitalization and death, that there may also be disparities in the likelihood of knowing a loved one who has suffers from a physical illness, including COVID-19.
Lines 62-64- Good mention of the distribution of stimulus checks. Also worth consideration is existence of any racial or ethnic disparities in receipt of stimulus checks due to format of check delivery (e.g., direct deposit versus check), access to Internet, and other factors, which could also affect mental health outcomes.
Lines 156-158- Do the government aids examined also include the Women, Infants and Children (WIC) program?
Line 160- For the collapsed age categories, the age range of 23-64 seems to be a wide range. Would this confound the results?
Lines 279-281- In the line "One previous study finds that the impact of unemployment is stronger for those with high social support or a low previous wage [33]", it is unclear if the word 'stronger' denotes a positive or negative effect. How is 'stronger' defined?
Lines 309-311 - Has the effect of social support been investigated in the role of more favorable mental health outcomes among the Black group?
Overall, this is an insightful, compelling and interesting paper.
Author Response
Comment: Lines 15- 17 and 167- With the sentence, "Our statistical 15 analysis uses the ordinal regression model (ordered logit model) that takes both the week and 16 regional factors into consideration to control for potential time effects", what effect would the week have on mental health? This is referenced also in line 167 but could be expanded upon.
Response: Thank you for pointing this out. We agree with your suggestion. According to CDC (https://www.cdc.gov/mentalhealth/learn/index.htm#:~:text=Yes%2C%20it's%20important%20to%20remember,mental%20health%20could%20be%20impacted.), a person’s mental health can change over time, depending on many factors. During the outbreak of a pandemic, like COVID-19, individuals face extra pressure that might exceed their coping abilities, their mental health could be impacted. Our observations from the COVID-19 Household Impact Survey were collected during the periods of April 20-26, May 4-10, and May 30-June 8, and the virus kept spreading in the U.S during these periods. Besides the dynamics of health risks due to COVID-19, an individual might have to worry about food or survival if he or she lost the job, which leads to poorer mental health. Your suggestion on more explanation about time effects has been addressed in line 180-183 of the “Statistical Analysis” section.
Comment: Lines 43-44 - Good discussion. It may be worth considering that since there are racial and ethnic disparities in COVID-19 sickness, hospitalization and death, that there may also be disparities in the likelihood of knowing a loved one who has suffered from a physical illness, including COVID-19.
Response: Yes, we agree with your point. Here we are trying to point out some specific mental stressors that have been examined in previous studies. Our paper primarily focuses on unemployment and financial aid, while the concern about one’s relatives’ health status is not our primary focus. We agree that these disparities do exist but adding or expanding on them may cause trivial diversions when laying down the rationale in the introduction, but it is worth considering the disparities in the next study.
Comment: Lines 62-64- Good mention of the distribution of stimulus checks. Also, worth consideration is existence of any racial or ethnic disparities in receipt of stimulus checks due to format of check delivery (e.g., direct deposit versus check), access to Internet, and other factors, which could also affect mental health outcomes.
Response: Yes, this is a very good recommendation. The racial or ethnic disparities in receipt of stimulus checks might have different impacts on mental health across races. And we thought about categorizing forms of receipt of the financial aid, but the survey, which is the source of our data, did not specify the format of assistance delivery. Therefore, we are unable to further differentiate the forms of receiving financial aids. But we could investigate it in the future research when there are more data about the financial aids program.
Comment: Lines 156-158- Do the government aids examined also include the Women, Infants and Children (WIC) program?
Response: According to the survey, these programs are not included in the aids.
Comment: Line 160- For the collapsed age categories, the age range of 23-64 seems to be a wide range. Would this confound the results?
Response: Thank you for pointing this out. Yes, admittedly, the age range of 23-64 is a wide range. However, in the case of our study, the survey data only report three categories for everyone, i.e. 18-22, 23-64, and 65+, so we have to adopt these categories. In addition, the Age is a control variable in this analysis, which could not affect the major points of our paper.
Comment: Lines 279-281- In the line "One previous study finds that the impact of unemployment is stronger for those with high social support or a low previous wage [33]", it is unclear if the word 'stronger' denotes a positive or negative effect. How is 'stronger' defined?
Response: Yes, it is a good observation and thanks for pointing this out. The line is changed to “One previous study finds that the adverse impact of unemployment on mental health is stronger for those with high social support or a low previous wage [30]” in line 286-288.
Comment: Lines 309-311 - Has the effect of social support been investigated in the role of more favorable mental health outcomes among the Black group?
Response: This is a good question. Regarding this more favorable mental health outcomes among the African American group, we have done extensive research on writings addressing this phenomenon. It is worth noticing that the reasons are still under debate, almost perplexing. Various empirical research specifically addressing religious participation, social groups or neighborhood interactions could not anchor upon a solid explanation. Therefore, we could point out the phenomenon here, but could not offer a robust answer for the cause. We made this acknowledgement in the conclusion part of the paper and pointed out that further research could specifically target on the urban non-urban and African American non- African American differences.
Comment: Overall, this is an insightful, compelling and interesting paper.
Response: Thank you very much!
Reviewer 2 Report
The manuscript "Impacts of Lay Offs and Government Assistance on Mental Health during COVID-19: An Evidence-Based Study on the United States” describes a study evaluating the impact of unemployment and government financial assistance during the COVID-19 pandemic on the working-age population's mental health.
The question posed by the authors is well defined in the introduction and appears to fit the aims and scope of the journal. The design of the study is clear and the methods are replicable. The statistical analysis are appropriate to the study design.
By examining the differential impact on urban and non-urban groups as well as black and non-black groups, the present study appears to significantly contribute to the research field. In particular, differently than in previous studies, the present paper shows that the urban population is psychologically more vulnerable to unemployment compared to their non-urban counterparts. Moreover, the present study demonstrates that the black group is more resilient and that government financial assistance does not ameliorate the mental health conditions of the working-age population.
In particular, I would suggest to the Authors to describe Tables 10-11-12-13 in the Results section and to fully dedicate the Discussion section to comments and comparisons with other data in the literature. Although difficult, I would invite the authors to expand the interpretative hypotheses relating to some of their most significant findings (resilience of black and non-urban groups). Moreover, the first part of the Discussion section (lines 258 to 277) appears to be more appropriate as conclusion of the paper (after discussing the results).
Finally, the authors should double-check the text for typing errors (e.g. line 20: one the other hand).
Author Response
Comment: The manuscript "Impacts of Lay Offs and Government Assistance on Mental Health during COVID-19: An Evidence-Based Study on the United States” describes a study evaluating the impact of unemployment and government financial assistance during the COVID-19 pandemic on the working-age population's mental health.
The question posed by the authors is well defined in the introduction and appears to fit the aims and scope of the journal. The design of the study is clear, and the methods are replicable. The statistical analysis is appropriate to the study design.
By examining the differential impact on urban and non-urban groups as well as black and non-black groups, the present study appears to significantly contribute to the research field. In particular, differently than in previous studies, the present paper shows that the urban population is psychologically more vulnerable to unemployment compared to their non-urban counterparts. Moreover, the present study demonstrates that the black group is more resilient, and that government financial assistance does not ameliorate the mental health conditions of the working-age population.
In particular, I would suggest to the Authors to describe Tables 10-11-12-13 in the Results section and to fully dedicate the Discussion section to comments and comparisons with other data in the literature.
Response: Thank you for your suggestion. We agree with this and have incorporated the description about Tables 10-11-12-13 in the Results section and dedicated the Discussion section to fully comments and comparison with other data in the literature.
Comment: Although difficult, I would invite the authors to expand the interpretative hypotheses relating to some of their most significant findings (resilience of black and non-urban groups).
Response: Yes, we have addressed this comment by amending the discussion on African American and non-urban groups. While the higher resilience in non-urban group could be expanded to the impacts of built environment (line 297-303), it is hard to thoroughly explain the higher resilience in the African American group as no research has fully explained the mechanisms behind. Deducing from the theoretical discussion from the literature review, we could expect that the higher resilience of the African American population is due to their closer social network therefore higher social capital within their social network and community (line 314-319). However, no literature has done robust research about the topic that could support this hypothesis. We have to acknowledge that future research is needed to address that.
Comment: Moreover, the first part of the Discussion section (lines 258 to 277) appears to be more appropriate as conclusion of the paper (after discussing the results).
Response: Yes, it is a good suggestion. We have modified the paper so that that paragraph is synthesized in the conclusion.
Comment: Finally, the authors should double-check the text for typing errors (e.g. line 20: one the other hand).
Response: Thank you for pointing out that. We will proofread through to avoid similar errors.
Reviewer 3 Report
Throughout the paper, critical facts and data about unemployment and government aid are not referenced.
Why are authors using "black group" as a variable in this study?
Who are "non-black groups" and why is this construct important to the question of government aid and mental health?
The social mechanisms potentially involved in the authors use of the construct black group, government aid, and mental health are not discussed so it is unclear why this research question is relevant?
The key concepts and descriptions of data collection are not described.
The assumptions, methods and resulting findings provide no benefit to the existing literature in psychology or economics.
I have major ethical concerns with the assumptions and logic presented in this paper.
Author Response
Comment: Throughout the paper, critical facts and data about unemployment and government aid are not referenced.
Response: Thank you for pointing that out. Please refer to the Introduction section, Lines 32-34 described the estimates of Bureau of Labor Statistics on unemployment in the U.S. (14.7% in April 2020), and lines 65-66 have the IRS’s data on total of stimulus checks ($159 million by June 24, 2020). In addition, our data is collected from the survey, which documented various types of government aids received by those surveyed respondents, but the actual number of how much they received is not included in the survey questionnaire. Also, please find the description of statistics about unemployment and government aid from the survey in Table 1-2-3.
Comment: Why are authors using "black group" as a variable in this study? Who are "non-black groups" and why is this construct important to the question of government aid and mental health?
Response: Thanks for pointing this out. Lines 90 – 93 explains this. Existing research found that the African American community in general enjoys more favorable mental health status and exhibit stronger resilience to disasters and we wanted to follow the logic of previous studies to investigate the mental well-being of African Americans during the abnormal times, such as the outbreak of COVID-19. To give more sensitivity in the writing, we changed ‘black’ into African American throughout our paper. In addition, we have done specific statistical tests for all other ethnical groups (White, Hispanic, Asian and Others, which are all categorized in the survey), and only found that the AA groups has significantly higher mental resilience, which corresponds to the findings of existing literature. Our literature research found multiple articles have observed this phenomenon in their empirical analyses, and we believe that this sensitivity should be given to our empirical analysis as well. The division between AA and non-AA groups are purely derived from existing research findings, and our statistical analysis confirms their findings on the stronger mental resilience exhibited by the African American community.
Comment: The social mechanisms potentially involved in the authors use of the construct black group, government aid, and mental health are not discussed so it is unclear why this research question is relevant?
Response: This question can be answered in two folds. Firstly, this paper is primarily interested in the effects of unemployment and government aids on the working-age population’s mental health. Reason for that is described in introduction (lines 29-36 and 71-77) that this paper is intended to examine the impacts of government interventions (aids) on the population mental health because few prior works were conducted to examine such impacts, particularly using data during COVID-19. The research findings could help policy makers better understand the actual outcomes of their interventions beyond the direct effect of economic relief. Reason for constructing AA and non-AA, urban and non-urban groups is that prior literature studies on similar topics found that the same stressors could have different impacts on these groups. To better present our empirical analyses and results, we divided the data into these groups to take care of the sensitivity of heterogeneity existing in the population. This division of data sample is purely derived for the appropriateness of statistical analyses.
Comment: The key concepts and descriptions of data collection are not described.
Response: Please refer to the Materials and Methods section of this paper, we have presented the details of data source and the descriptions of data collection here. More specifically, there are two methods of sampling designed by the NORC at the University of Chicago: AmeriSpeak (a probability-based panel designed to be representative of the U.S. household population) and ABS (Multi-mode Address Based Sample, which is a sampling frame based on an extract of the U.S. Postal Service delivery-sequence file). Please go to the official website (https://www.covid-impact.org/results) and check the field reports if you want to know the very details of both sampling techniques. The key concepts, including mental health and 4 major independent variables are introduced in Materials and Methods section, and the descriptions of them are presented in Table 1-2-3.
Comment: The assumptions, methods and resulting findings provide no benefit to the existing literature in psychology or economics.
Response: First of all, our paper contributes to the existing literature in both economics and psychology by evaluating the outcome of government financial assistance during COVID-19 on the population’s mental health. Our paper, based on existing literature, made the hypothesis that layoff could affect the population’s mental health through both the financial and the psychological pathways. Government assistance, although designed to address the financial pathway, could not address the psychological pathway therefore its actual outcome awaits an empirical analysis (Lines 101-103). Our empirical analyses support our hypotheses and made the conclusion that financial aids could adversely affect the population’s mental health. This not only fills research gap, but also could inspire policy makers.
This paper further examines the differential impacts on urban and non-urban groups, as well as AA and non-AA groups based on the findings of the existing literature. By examining the differential impacts on urban and non-urban groups as well as AA and non-AA groups, our paper contributes to the research in psychology since no other research has given similar attention and our paper found that the urban population is psychologically more vulnerable to unemployment compared to their non-urban counterparts, while confirming with previous research that the AA community is more resilient.
Comment: I have major ethical concerns with the assumptions and logic presented in this paper.
Response: We are sorry that our paper could not present itself without confusion at your first glance. Per your questions, we have: made clear the facts about unemployment and government aids, explained the reason for making the AA group as a variable, made clear the research questions and data collection, and highlighted the contribution to the existing literature in economics and psychology. Hope through these efforts, you could now have a refreshed understanding of our paper and all the confusions are successfully addressed.
