![ORCID Icon]({"type":"image","src":"/templates/jsp/_style2/_tandf/images/orcid.svg"} "ORCID Icon"){kind=small}
https://orcid.org/0000-0002-6836-3030
ABSTRACT
ABSTRACT
This study examined the effect of socio-economic features of low-income communities and COVID-19 related cases in New York City. The study developed hypotheses and conceptual framework of low-income communities and COVID-19 associated cases based on literature and theoretical review. The proposed framework was then tested using Structural Equation Model (SEM) with secondary data collected from New York Health and Mental Hygiene Department, US Census Bureau, and the Centers for Disease Control and Prevention. The findings revealed that unfavourable working conditions, underlying health conditions, and poor living conditions significantly and positively affects the number of COVID-19 confirmed cases. The study further revealed a positive and significant relationship between confirmed COVID-19 cases and COVID-19 related deaths. Theoretically, this study provides empirical results and a conceptual framework that could be used by other researchers to investigate low-income communities and COVID-19 related topics. Practically, this study called on the federal and state governments to effectively apply the health justice approach to eliminate healthcare discrimination for people living in low-income and marginalised communities as well as providing accessible, safe housing for the more vulnerable who need a place to self-quarantine due to COVID-19 exposure. Further practical and theoretical implications policies are discussed.
Introduction
In March 2020, WHO declared the Coronavirus disease 2019 (COVID-19) outbreak as a global pandemic after its first outbreaks in Wuhan, China, in December 2019 (Ciotti et al., 2020). Since then, the virus has rapidly infected more than 13 million people; and taken the lives of more than 500,000 people worldwide as of Jun 23th, 2020 (Worldometer, 2020).
The US leads the countries that are heavily affected by the pandemic. The country accounts for nearly 27% of confirmed cases worldwide, with more than 2 million confirmed cases and over 115,000 reported deaths as of June 23th, 2020, (CDC, 2020a; John, 2020; Worldometer, 2020). In contrast, much attention on the spread of the virus has focused on the vulnerability of seniors and Americans living with underlying health conditions such as hypertension, diabetes, asthma, and other chronic diseases. Public health experts and leaders have contended that unfavourable social and economic conditions such as racial, financial, and geographical location of low-income communities serve a greater risk of contracting the virus(Amin, 2020; Szabo & Recht, 2020).
Despite calls from public health experts and community leaders to governments and policymakers to aid low-income communities during these challenging times, there is little to no empirical studies to ascertain the contributing factors of low-income communities and the affected cases of COVID-19. Hence, it is imperative to understand the relationship between the number of people living in low-income communities and the confirmed COVID-19 related cases.
The current study aims to bridge the gap by empirically examining the relationship between socio-economic features of low-income communities and COVID-19 related cases. The present study is among the first research to unearth the relationship between socio-economic features of low-income communities and confirmed COVID-19 Cases in North America. The study also contributes to existing literature and serves as a benchmark for future COVID-19 related research. Furthermore, this study will help guide governments, public health experts, and policymakers to make informed decisions in the quest to eradicate the virus.
The rest of the paper is organised as follows. The next section delves into the literature review, research methodology, which includes the data collection and modelling process used to test the hypothesis and the conceptual framework. The analysis and empirical results are then presented, followed by theoretical and practical discussion. The final section concludes the research findings with discussion and policy recommendations.
Literature review
The rapid increase in COVID-19 cases in the US raises major concerns about the nation's economic inequalities and fragile social safety net that might put the vulnerable groups into the ‘economic brunt of the crisis’ (HRW, 2020).
While the virus does not discriminate against people regardless of their race, and wealth, many social and healthcare experts have cautioned that low-income communities are the most vulnerable in contracting the virus due to the decrease of economic mobility and the high cost of medical care, (Amin, 2020; Bolin & Kurtz, 2018; CDC, 2020b; Green-Laughlin, 2020; HRW, 2020).
The city of New York in the United States has reported high rate cases of COVID-19 fatalities in racialized and low income Hispanic and Black communities, which accounts for more than 62 percent of the related deaths in the state (Wilson, 2020).
A report by the US Center for Disease Control CDC (CDC, 2020b; Szabo & Recht, 2020) revealed that, the types of work and policies within a working environment where people in some racial and ethnic groups are overrepresented, are more likely to contract the virus. They further mentioned that low-income communities’ infection rate could be higher for workers in the essential industries who continue to work outside despite the outbreak within their communities due to their language barrier, low economic and living conditions.
Furthermore, New York is counted among states such as California and Washington state, where over seven million people face language barriers with more than 200 spoken language barriers (Courts, 2020). Evidently, people with English language barriers might be at a greater risk of contracting the virus since most of them are compelled to work as janitors, housekeepers, in hotels and hospitals where they are usually exposed to work hazards (Statistics, 2015; Velasquez et al., 2020).
In addition, employees with limited English language proficiency do not receive adequate health information from their employers and the community (Jackson & Fitzsimons, 2017). This information gap could worsen health disparities since most of these people do not have the adequate information needed to take preventive measures at the workplace and in their communities to protect themselves from getting infected by the virus (Alegria, 2020).
Moreover, people with English language barriers are more likely to live in crowded communities with close living conditions and poor infrastructure, which exposes them to the virus and other health conditions (Velasquez et al., 2020).
Low-income African Americans and Hispanic communities in the US continue to experience stigma and systematic health inequalities such as not having health insurance, and other challenges in getting access to healthcare (Bravo et al., 2019; Cunningham et al., 2017; Hearst et al., 2008). These factors make the population of any racial and ethnic minority groups more vulnerable and exposed to many challenges in public health emergencies such as the COVID-19 outbreak. Undeniably, these factors contribute to a higher rate of low-income communities contracting the virus, (Bolin & Kurtz, 2018; Bravo et al., 2019; CDC, 2020b; Cunningham et al., 2017). Correspondingly, research conducted by (Bor et al., 2017) indicated a strong association between low income and poor health conditions. (CDC, 2020c) reported that people with health conditions such as Chronic kidney disease, COPD (Chronic obstructive pulmonary disease), Obesity, Coronary heart disease, Diabetes, High blood pressure are more at risk of severe illness from the virus.
Furthermore, poor living conditions of people in low-income communities such as living in densely populated areas, congregate living, residential segregation, distancing away from essential facilities, and multi-generational households serve as major contributors to underlying health issues. These conditions make it difficult for people living in these communities to follow preventive measures during public emergencies and to seek medical treatments, especially during the COVID-19 outbreak, (Bartel et al., 2019; Bolin & Kurtz, 2018; Bravo et al., 2019; CDC, 2020b; Hawkins et al., 2019; HRW, 2020; Sammy, n.d.; Szabo & Recht, 2020; Wilson, 2020).
Based on the preceding theoretical and empirical discussion, the following hypotheses are developed.
H1: There is a positive and significant relationship between people with unfavorable working conditions and the number of COVID-19 Confirmed Cases.
H2: There is a positive and significant relationship between people with underlying health issues and the number of COVID-19 Confirmed Cases.
H3: There is a positive and significant relationship between people with poor living conditions in low-income communities and the number of COVID-19 Confirmed Cases
H4: There is a positive and significant relationship between the COVID-19 Confirmed Cases and the death rate of COVID-19.
Methodology
To examine the relationship between people living in low-income communities and the high rate of infected COVID-19 cases, this study analyzed the relationship between COVID-19 cases and low-income communities per zip code from February 29 to June 23, 2020, in New York City.
Data were collected from the New York Health and Mental Hygiene Department (NYC Health, 2020) with confirmed number of positive cases, death cases, and tests given per zip code, with population and median income estimated data being collected from the US Census Bureau (US Census Bureau, 2018).
Additional data on living and working conditions were gathered from the ATSDR (Agency for Toxic Substances and Disease Registry, 2016). The researchers further gathered data concerning underlying health conditions from the Centers for Disease Control and Prevention (CDC, 2017). All files were merged using Zip Code as a primary key. Since some data were reported at the census tract level, hence, census tract level datasets were allocated to zip code level using the HUD’s Office of Policy Development and Research ZIP Code Crosswalk (US Department of Housing and Urban Development, 2014).
Although the number of COVID-19 fatalities are disproportionately higher among US cities such as Florida, Texas, Chicago, Mississippi, Detroit, and Memphis, New York City was selected for the subject of the study since in March, 2020 the city was the epicenter of the virus in North America with a large number of confirmed cases. As of March 26, 2020, nearly 50% of confirmed cases in the US were confirmed in New York City. Since June 1, 2020, the city has the highest COVID-19 cases across America (CDC, 2020b).
New York City's low-income communities appear to be one of the hardest-hit cities in the US by the virus, with more than double the rate of positive and death cases compared to wealthy communities, according to data released by the city's health department (Edomt, 2020; Mansoor, 2020). Besides, New Yorkers are among disproportionately high-income communities in the US with people of color, especially African Americans and Latinos. Though blacks are only 22% of New York City's population, they constituted 66% of confirmed cases, with 68% of the death toll as of June 14 (Edomt, 2020).
Further, the New York Health Department is among the first institutions to collect data points per zip code starting on February 29, 2020. This helped the researchers to identify areas that are heavily affected by the virus.
In addition, the New York State’s income disparity is the highest in the nation. According to a report by the Economic Policy Institute (Cropley, 2018), the US's personal income disparity is getting significantly wider. The New York State is leading the country in this inequality, with the ratio between the average top income earners with the bottom ones as 44.4–1, compared to the countrywide numbers where the ratio is 26.3–1. This corresponds to recent research conducted by Clarke, H. and Whitely, P (Clarke & Whiteley, 2020), which revealed that states with greater income inequality are more likely to have higher cases of COVID-19 and related deaths.
After an extensive literature review, the critical factors that affect the number of COVID-19 cases in low-income communities in New York State were observed as detailed in Table 1, the conditions being Unfavourable Work Conditions (UWC), Underlying Health Conditions, (UCH), Living Condition (PLC), Low Income (MI), Confirmed Cases (CC), Death Rate (DR).
Table 1. List of factors affecting the number of COVID-19 cases in low-income communities in New York State.
Data analysis
The techniques used for the data analysis in this study are comparable or partially comparable to those of earlier studies (Bartsch et al., 2020; Gbongli et al., 2019; Xiao et al., 2020) to validate the research framework and test the suggested research hypotheses.
Using SmartPLS, a PLS-SEM was employed to explore the relationship between independent and dependent variables by examining path analysis, confirmatory factor analysis, second-order factor analysis, regression models, covariance structure models, and correlation structure models.
Partial Least Squares Structural Equation Modeling (PLS-SEM) has become a popular tool that is widely used in social science studies, to estimate complex relationship and models with many constructs, indicator variables, and structural paths without requiring distributional assumptions on the data (Hair et al., 2019; Sarstedt & Cheah, 2019). Evidently, PLS-SEM is currently the most appropriate technique used in social sciences studies; as such, applying this technique was the appropriate method for the multivariate analysis in this study.
Evaluations of outer measurement model
As recommended by Henseler et al. (Henseler et al., 2009), PLS includes a two-step process that requires evaluating the outer measurement model and evaluating the inner structural model.
Considerably, the outer measurement model was used to calculate the reliability, internal consistency, and validity of the observed variables along with the unobserved variables (Ho, 2006). Outer loadings of all items met the threshold ranging from 0.483–0.981, as shown in Table 2. Cronbach's alpha and Composite Reliability (CR) were used to analyze the construct reliability's internal consistency evaluation. The higher the value indicates, the higher the reliability levels (Hair et al., 2019). Table 2 shows the Cronbach's alpha and Composite Reliability for all reasonably reliable variables as all constructs were greater than the threshold of 0.6. Average Variance Extracted (AVE) was calculated to address the convergent validity of each measured construct. As indicated in Table 2, the results indicated that all measured constructs in the study met the threshold of 0.5 (Hair et al., 2019); therefore, convergent validity was confirmed for the research model.
Table 2. Construct Reliability and Validity.
Discriminant validity
Fornell-Larcker criterion was used to analyze the discriminant validity of the latent constructs to examine the distinctiveness of the manifest variable from other constructs in the path model. Discriminant Validity helped the researchers ascertain whether the model's constructs are highly correlated with each other or not by comparing the Square Root of AVE and correlation from other latent constructs.
As indicated in Table 3, discriminant validity in this study indicated that our dependent variable was sufficiently different or distinctive from other constructs given the fact that all correlations were smaller compared to the Squared Root of AVE. Therefore, the suggested conceptual model was accepted, with the confirmation of adequate reliability, convergent validity, discriminant validity, and the research model's substantiation.
Table 3. Fornell-Larcker Criterion Test.
Evaluations of the inner structural model
After confirming the validity and reliability of the measurement model, the researchers further measured the Inner Structural Model outcomes using the coefficient of determination (R2), Effect size (ƒ2), Goodness-of-Fit (GOF) index, Correlation coefficient of latent variables, Path coefficient (b value), and T-statistic value. Table 4.
Table 4. Cross- Loadings.
Measuring the value of R2
In the quest to measure the predictive model's accuracy, the coefficient determination was used to measure the total effect size, and Variance explained in the dependent variables. The inner path model of the study was 0.581 and 0.565. This indicates that the three independent constructs substantially explain 58.1% of the Variance in the confirmed COVID-19 cases, while the confirmed COVID-19 rates explain 56.5% of the Variance in death cases. R2 value ranges from 0 to 1, with the higher the value, the greater the predictive power. R2 values of 0.75, 0.5, and 0.25 can be considered substantial, moderate, and weak (Hair et al., 2019; Henseler et al., 2009). Therefore, the R2 value in this study is moderate. Table 4.
Measuring the size effect
The ƒ2 explains the degree of the impact of each independent construct on the dependent construct. The value of a coefficient determination changes whenever an independent construct is deleted from the path model. This explains if the deleted or removed independent construct has a significant effect on the dependent variables. The threshold values for ƒ2 are 0.35 (strong effect), 0.15 (moderate effect), and 0.02 (weak effect) as defined by (Cohen, 1988). Table 5 below demonstrates that all four exogenous latent constructs’ effect size had a strong total effect from the calculation.
Table 5. Size Effect.
Goodness of fit-index
Goodness-of-Fit (GOF) was applied as an index for the complete model fit to verify that the model sufficiently explains the empirical data (Tenenhaus et al., 2005). The Goodness of Fit Indices (GOF) values lies between 0 and 1, where values of 0.10 (small), 0.25 (medium), and 0.36 (large) indicate the global validation of the path model. As indicated in Table 6 below, following the recommendation of earlier works (Gbongli et al., 2019; Tenenhaus et al., 2005; Wetzels et al., 2009), the GOF index for this study model was measured as 0.65 by employing the Average Variance Extracted (AVE) = 0.7406 and the average R2 values = 0.573 (See Table 7 below). Therefore, the empirical data fits the model satisfactorily and has substantial predictive power in comparison with baseline values.
Table 6. Goodness-of-Fit index calculation.
Table 7. Model fit summary.
The SRMR is a measure of the estimated model fit. When SRMR =< 0.08, the study model has a good fit (Hu & Bentler, 1998), with a lower SRMR a better fit. Table 7 below reveals that the study model had a moderate SRMR result of 0.182.
However, for the purpose of this study, Absolute Fit Indices was used as the basis for the Goodness of Fit Indices to measure how well our theoretical model fits the observed data collected, (Hussain et al., 2018). The statistical-based measure of the Absolute Fit used was the Chi-square (CMIN) statistic and its associated p-values. Statistically, all p-values were significant from 0.000–0.010 and a Chi-square (CMIN) of 3122.472 > 3, which implies the acceptance of the absolute fit indices for this study. According to (Taufique, 2015), ‘Chi-square (CMIN) statistic and its associated ‘probability’ or p-value should not be statistically significant if there is a good model fit.’ NFI was also was measured by computing the Chi² value of the proposed model (Bentler & Bonett, 1980). The study's NFI was 0.519, which met the threshold of 0 and 1 representing an acceptable fit (Lohmöller, 1989).
Correlation coefficient of latent variables
Table 8 below shows the latent variable correlation coefficient, which indicates a strong correlation between the latent exogenous constructs and the latent endogenous construct.
Table 8. Latent Variable Correlation.
Table 9. Path coefficient and T-statistics.
Estimation of path coefficients and T-statistics
The study hypothesis's significance was tested using β values: the greater the β value, the more the substantial effect on the manifest latent construct. Yet, the β value had to be verified for its significance level through the T-statistics test. Bootstrapping technique was applied to examine the significant path coefficient and T-statistics of the research hypothesis using 5000 subsamples and an alpha value of .05 in Smart-PLS.
After an extensive literature review, the researchers predicted that unfavourable working conditions would significantly and positively influence the confirmed rates of COVID-19 (H1).
The findings in the above table confirmed this hypothesis (β = 0.211, T = 2.695, p = 0.007); thus, H1 was accepted. In addition, the researchers depicted (H2) positive and significant relationships between people with underlying health issues and the high number of infected COVID-19 Cases. The research findings above confirm this hypothesis with (β = 0.221, T = 2.586, p = 0.010). Furthermore, the analysis revealed a positive and significant relationship between people with poor living conditions and the infection rate of COVID-19 with (β = −0.665, T = 11.365, p = 0.000), which confirms (H3). Similarly, the confirmed rate has a significant and positive impact on the death rate with (β = 0.752, T = 24.461, p = 0.000), confirming (H4). The following figure shows the graphical representation of all path coefficients of the model. Figures 1–3.
Published online:
21 November 2020![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/rgph20/2021/rgph20.v016.i01/17441692.2020.1850830/20210521/images/medium/rgph_a_1850830_f0001_oc.jpg"})
Figure 1. Proposed Research Framework.
Published online:
21 November 2020![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/rgph20/2021/rgph20.v016.i01/17441692.2020.1850830/20210521/images/medium/rgph_a_1850830_f0002_oc.jpg"})
Figure 2. Assessment of the structural equation model.
Published online:
21 November 2020![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/rgph20/2021/rgph20.v016.i01/17441692.2020.1850830/20210521/images/medium/rgph_a_1850830_f0003_oc.jpg"})
Figure 3. Graphical representation of the path coefficient.
Discussion
The critical contribution of this research paper is to identify the socio-economic and health factors that lead to the number of confirmed rates of COVID-19 in New York City by using Pearson correlation and PLS-SEM techniques.
These techniques are effective measurement in determining and analyzing the complex relationship between multiple variables. The study results reveal that working conditions, underlying health conditions, and living conditions in low-income communities have a positive and moderate effect on the confirmed COVID rates (R2 = 0.581). Similarly, the confirmed COVID-19 related cases also positively and moderately affected the death rate (R2 = 0.565, p = 0.000).
Moreover, the study provided a substantial GOF (GOF = 0.65). The final SEM results revealed that underlying health conditions have the highest path coefficient (β = 0.221), with the overall leading to the number of confirmed rates. The underlying health conditions were found to be the second utmost factor (β = 0.211) of the overall factors leading to COVID-19 rates. Thus, a workplace operating during the pandemic needs to have a protocol in place to protect its employers from contracting the virus. All socio-economic and health factors relating to the confirmed rates were found to substantially affect the confirmed rates.
In addition, the confirmed rates show a strong effect on the death rates where the beta is 0.752, and the p-value equals 0.000. The findings of this paper revealed that all suggested hypotheses are supported by a positive and statistically significant relationship. Thus, all the hypotheses presented in this study were accepted.
The relationship between all the conditions identified in this study and COVID-19 related cases suggests that vulnerable communities in New York City are at higher risk of contracting the virus, resulting in higher death rates in these areas. People who continue to work during this challenging time, such as retail workers, factory employees, farmworkers, and healthcare staff, are faced with threats of being affected. This result is in line with the findings of earlier reports. The Washington Post (Bhattarai, 2020) reported that 41 retail workers across the country have died from the virus and with more than 1,500 confirmed COVID-19 cases and unconfirmed cases of 3,000 workers who are either quarantined, hospitalised, or waiting for their test results. For healthcare professionals, the number is significantly higher, with more than 97,000 confirmed cases, along with nearly 520 deaths, according to the Centers for Disease Control and Prevention (CDC, 2020a), with the majority of them being people of color and immigrants (McNicholas & Poydock, 2020; Rho et al., 2020; Travis Du Bry, 2014).
Thus, besides providing protective gear and proper cleaning policy in the workplace, it is recommended that employers should provide education and training about COVID-19 preventive measures such as sanitisation and social distancing before the start of each shift.
In addition, due to the problem of low English language proficiency, employers should provide multilingual booklets that contain details about COVID-19 and its preventive measures to their employees and encourage them to go through those materials before each shift to ensure that all employees are on the same page in helping fight the virus.
Moreover, the US is among one of the countries that do not provide paid sick leave from the very first day of sickness (Heymann et al., 2020), with 54% of Latinx workers and 38% of black workers not qualified for paid sick leave (National Partnership for Women & Families, n.d.). Therefore, paid sick leave due to Covid-19 infection and exposure should be made available for all the workers.
Moreover, asymptomatic employees should be encouraged to work from home whiles they are paid if possible. This is due to the fact that people who are asymptomatic and not aware that they have the virus are more likely to spread the virus at the workplace and in their communities due to their activeness and going about their daily activities.
Additionally, the positive relationship between underlying health conditions and confirmed rates reveals that those with poor health conditions face higher risks of contracting the virus. This finding confirms the study by Andrew Clark et al. (Clark et al., 2020), which reveals that those with underlying health conditions can be at risk of a severe case of COVID-19 if infected, varies with age where the older the patients, the higher the risk. Studies conducted by (Bartel et al., 2019; Bolin & Kurtz, 2018; Bravo et al., 2019; CDC, 2020b; Hawkins et al., 2019; HRW, 2020; Sammy, n.d.; Szabo & Recht, 2020; University of Aberdeen, 2020; Wilson, 2020) show that underlying health conditions might subsequently lead to higher COVID-19 death risk. For that reason, it is recommended that persons who are at higher risk need to shield themselves with proper protective gears and more intensive physical distance. People with underlying health conditions usually have a routine schedule for getting health care from hospitals or their family physician. It is imperative for healthcare policymakers to provide in house health services to such persons with underlying health conditions to avoid exposing them to the virus at healthcare facilities or within their community. Healthcare workers who would be assigned to persons with underlying health conditions should be well educated, trained, and informed about all COVID-19 preventive measures for providing in-house or remote healthcare services.
Further, there should be continuous public health education on the importance of exercising, diet control, following medical protocols from health experts, from the media, and other organisations, especially for people with underlying health issues, to boost their immune system to help reduce the spread and infectious rate. With the support of physicians and nurses, healthcare policymakers could create and provide a daily exercise, and eating plan made publicly available to all, particularly those with underlying health conditions, during these lockdown periods.
Finally, our findings also revealed that poor living conditions are statistically significant with COVID-19 related cases since it is difficult for people in crowded households or communities to practice physical distancing or to observe public health emergency protocols. Moreover, people who do not own vehicles will result in use of public transit, which increases their risk of being affected. These findings are in line with the previous studies of (Bartel et al., 2019; Bolin & Kurtz, 2018; Bravo et al., 2019; Centers for Disease Control and Prevention, 2020; Hawkins et al., 2019; HRW, 2020; Sammy, n.d.; Szabo & Recht, 2020; Wilson, 2020). The findings of this study is also uniform with research conducted by Northwell Health's Feinstein Institutes for Medical Research (Misra, 2020). Thus, it is critical to provide safe housing for quarantine and sufficient cleaning supplies for people living in crowded communities with poor infrastructure. Therefore, this study recommends that employers offer transportation services fully operated by their firm to pick up and drop off employees to and from work. By so doing, firms would ensure that their vehicles are well cleaned and sanitised with preventive and social distance measures in place for passengers (employees) to help stop the spread of the virus.
Theoretical and practical recommendations
Theoretical recommendation
In terms of theoretical recommendation, the research paper contributes to current knowledge of the vulnerable communities and COVID-19 related cases. Several studies have identified that low-income communities, communities of color, people with disability and the elderly have a disproportionately high number of COVID-19 related cases (Amin, 2020; Bartel et al., 2019; CDC, 2020b; Millett et al., 2020). Nevertheless, there is little to no research on providing the holistic views on socio-economic features of low-income communities and COVID-19 related cases. Therefore, with this paper, we aimed to expand the current knowledge using the Structural Equation Modeling (SEM) techniques on this essential subject and to provide a benchmark for future research.
Practical recommendation
Our study reveals that poor living conditions, unfavourable working conditions, and underlying health conditions pose as the greatest risk of contracting the COVID-19 virus.
Accordingly, health protocols such as keeping social distance, practicing adequate sanitation at the workplace, and wearing protective gear are crucial in reducing the virus's spread and infection by workers. As mentioned in the discussion, employers should provide a multilingual COVID-19 preventive measures handbook and training to all employees to better keep employees on the same page in the quest of fighting the virus.
More importantly, paid sick leave and eviction freeze should be made available to all workers regardless of their work status to ensure those who show symptoms of the virus can stay home to protect themselves and their coworkers.
Moreover, asymptomatic people should be encouraged to work from home through teleworking, if possible, to help stop spread the virus at their workplaces and in their communities. There is also the need for the federal and state government to effectively apply the health justice approach to eliminate healthcare discrimination for people living in low-income and marginalised communities for healthcare services, such as testing and treatment for COVID-19. Health facilities with support from the federal and state government, businesses, and individual bodies should make cleaning and sanitation supplies largely available within low-income, marginalised, and vulnerable COVID-19 communities to help stop the spread.
Considerably, people with underlying health conditions should be educated and encouraged to shield themselves with better PPEs such as hand sanitisers, masks, and gloves when outside or in crowded areas to reduce their risks of infection. Additionally, public health agencies and the media should continue to emphasize the importance of healthy eating, physical distancing, and adequate hygiene practice to help prevent the spread of the virus and the fatality rate.
Lastly, continuous federal and state financial support is essential for businesses and property owners who continue to operate within low-income communities to reduce production costs to respond to the pandemic. The Families First Coronavirus Response Act passed by the US House of Representatives to provide several appropriations, such as guaranteed free coronavirus testing, paid leave and unemployment insurance, and enhanced benefit for children and families is a right step in the right direction which needs to be commended, (Appropriations, 2020).
However, since the bill did not address housing or homelessness issues, it is imperative for the federal and state government to provide accessible, safe housing for the homeless, more vulnerable, and those who need a place to self-quarantine due to COVID-19 exposure.
Conclusion
The COVID-19 pandemic has spread rapidly worldwide, which has affected more than 7.5 million people and taken the lives of over 400,000 people. The US is among the hard-hit countries globally, which accounts for nearly 26% of confirmed cases.
Many health experts have raised worries regarding vulnerable groups being more likely to be heavily impacted by this unprecedented pandemic. The paper aimed to identify the relationship between COVID-19 cases with several socio-economic and health factors using Pearson correlations and PLS-SEM techniques.
The study confirms that people living in low-income communities with unfavourable working conditions, underlying health conditions, and poor living conditions are more likely to contract the virus. Also, the study confirms a positive and significant relationship between confirmed COVID-19 cases and fatality rates.
The study findings correspond with previous research that discussed different factors within low-income communities that affect the number of COVID-19 cases and fatality rates. Thus, the results of this study will help guide decision-making by policymakers, governments and healthcare experts to amplify their understandings of the vulnerable communities that are affected by the virus.
As the government rapidly increases the numbers of tests available and provides a universal test while this paper is being written, the study merely provides a reference on the relation between the studied factors and COVID-19 related cases during the mentioned amount of time.
| Code | Factors |
|---|---|
| Work Conditions (UWC), (Bolin & Kurtz, 2018; Bravo et al., 2019; CDC, 2020b; Cunningham et al., 2017; Hearst et al., 2008) | |
| UWC_1 UWC_2 UWC_3 UWC_4 | Unemployed due to Disability Issues English Language Issues Single parent household with children under 18 Minority (all persons except White, non-Hispanic) |
| Underlying Health Conditions (UCH), (CDC, 2020c) | |
| UHC_1 UHC_2 UHC_3 UHC_4 UHC_5 UHC_6 | Underlying Chronic Kidney Disease Underlying Coronary Heart Disease Underlying Diabetes Underlying High Blood Pressure Underlying Obesity Underlying Chronic Obstructive Pulmonary Disease (COPD) |
| Poor Living Condition (PLC), (Bartel et al., 2019; Bolin & Kurtz, 2018; Bravo et al., 2019; CDC, 2020b; Hawkins et al., 2019; HRW, 2020; Sammy, n.d.; Szabo & Recht, 2020; Wilson, 2020) | |
| PPLC_1 PPLC_2 PPLC_3 PPLC_4 PPLC_5 | (Occupied housing units), more people than room capacity. Persons in institutionalised group quarters Housing in structures with ten or more units Households with no vehicle available Median Income |
CC_1 CC_2 CC_3 | COVID-19 Confirmed Cases (CC) (CDC, 2020a) Rates of positive COVID-19 cases per 100,000 People by ZCTA Percentage of people ever tested for COVID-19 with a polymerase chain reaction (PCR) test which tested positive Rates of hospital admitted COVID-19 patients tested positive with influenza and pneumonia like symptoms. |
DR_1 DR_2 | Death Rate (DR) (CDC, 2020a) Count of confirmed deaths Rate of confirmed deaths per 100,000 people by ZCTA |
| Main Construct | Code | Loading | Cronbach’s Alpha | CR | AVE |
|---|---|---|---|---|---|
| Unfavourable Work Conditions (UWC) | UWC_1 UWC_2 UWC_3 UWC_4 | 0.776 0.544 0.88 0.908 | 0.793 | 0.865 | 0.624 |
| Underlying Health Conditions (UCH) | UHC_1 UHC_2 UHC_3 UHC_4 UHC_5 UCH_6 | 0.981 0.941 0.969 0.965 0.924 0.967 | 0.983 | 0.985 | 0.918 |
| Poor Living Condition (PLC) | PLC_1 PLC_2 PLC_3 PLC_4 PLC_5 | 0.582 0.444 0.91 0.892 0.681 | 0.779 | 0.838 | 0.525 |
| Confirmed Cases (CC) | CC_1 CC_2 CC_3 | 0.934 0.92 0.824 | 0.873 | 0.923 | 0.799 |
| Death Rate (DR) | DR_1 DR_2 | 0.888 0.942 | 0.810 | 0.911 | 0.837 |
| CC | DR | PLC | UHC | UWC | |
|---|---|---|---|---|---|
| Confirmed Cases (CC) | 0.894 | ||||
| Health Rate (DR) | 0.752 | 0.915 | |||
| Poor Living Conditions (PLC) | −0.641 | −0.493 | 0.724 | ||
| Underlying Health Conditions (UHC) | 0.33 | 0.189 | 0.103 | 0.958 | |
| Unfavourable Working Conditions (UWC) | 0.392 | 0.311 | 0.008 | 0.845 | 0.79 |
| Code | UWC | UHC | PLC | DR | CC |
|---|---|---|---|---|---|
| UWC_1 | 0.776 | 0.822 | 0.171 | 0.161 | 0.187 |
| UWC_2 | 0.544 | 0.472 | −0.106 | 0.28 | 0.191 |
| UWC_3 | 0.88 | 0.647 | −0.098 | 0.323 | 0.386 |
| UWC_4 | 0.908 | 0.776 | 0.089 | 0.216 | 0.382 |
| UHC_1 | 0.836 | 0.981 | 0.227 | 0.146 | 0.212 |
| UHC_2 | 0.698 | 0.941 | 0.161 | 0.137 | 0.206 |
| UHC_3 | 0.869 | 0.969 | −0.035 | 0.283 | 0.424 |
| UHC_4 | 0.77 | 0.965 | 0.282 | 0.074 | 0.192 |
| UHC_5 | 0.839 | 0.924 | 0.092 | 0.154 | 0.381 |
| UHC_6 | 0.761 | 0.967 | 0.048 | 0.196 | 0.306 |
| PLC_1 | 0.544 | 0.461 | 0.582 | −0.004 | −0.094 |
| PLC_2 | 0.192 | 0.242 | 0.444 | −0.098 | −0.206 |
| PLC_3 | 0.254 | 0.253 | 0.91 | −0.306 | −0.491 |
| PLC_4 | 0.301 | 0.279 | 0.892 | −0.31 | −0.49 |
| PLC_5 | −0.551 | −0.318 | 0.681 | −0.625 | −0.636 |
| DR_1 | 0.233 | 0.09 | −0.441 | 0.888 | 0.569 |
| DR_2 | 0.324 | 0.235 | −0.461 | 0.942 | 0.78 |
| CC_1 | 0.332 | 0.307 | −0.575 | 0.671 | 0.934 |
| CC_2 | 0.389 | 0.335 | −0.671 | 0.59 | 0.92 |
| CC_3 | 0.327 | 0.24 | −0.466 | 0.757 | 0.824 |
| Confirmed Cases (CC) | Death Rate (DR) | Total Effect | |
|---|---|---|---|
| Confirmed Cases (CC) | 1.298 | Strong | |
| Death Rate (DR) | Strong | ||
| Living Conditions (PLC) | 1.023 | Strong | |
| Underlying Health Conditions (UHC) | 0.032 | Strong | |
| Working Conditions (UWC) | 0.03 | Strong |
| Average Variance Extracted (AVE) | R2 | |
|---|---|---|
| Confirmed Cases (CC) | 0.799 | 0.581 |
| Death Rate (DR) | 0.837 | 0.565 |
| Living Conditions (PLC) | 0.525 | |
| Underlying Health Conditions (UHC) | 0.918 | |
| Working Conditions (UWC) | 0.624 | |
| Average Values | 0.7406 | 0.573 |
| AVE * R2 | 0.424364 | |
| GOF | 0.651432 |
| Saturated Model | Estimated Model | |
|---|---|---|
| SRMR | 0.182 | 0.182 |
| d_ULS | 6.93 | 6.98 |
| d_G | 5.604 | 5.61 |
| Chi-Square | 3122.472 | 3124.487 |
| NFI | 0.519 | 0.518 |
| DR | UWC | UHC | PLC | LI | CR | |
|---|---|---|---|---|---|---|
| Death Rate | 1.00 | |||||
| Working Conditions | 0.323 | 1.00 | ||||
| Underlying Health Conditions | 0.238 | 0.843 | 1.00 | |||
| Living Conditions | −0.273 | 0.319 | 0.318 | 1.00 | ||
| Low-Income Communities | −0.57 | −0.537 | −0.307 | 0.318 | 1.00 | |
| Confirmed Rate | 0.677 | 0.379 | 0.334 | −0.502 | −0.585 | 1.00 |
| Standardize Beta | T Statistics | P Values | |
|---|---|---|---|
| Confirmed Cases -> Death Rate | 0.752 | 24.461 | 0.000 |
| Poor Living Conditions -> COVID-19 Confirmed Cases | −0.665 | 11.365 | 0.000 |
| Underlying Health Conditions -> COVID-19 Confirmed Cases | 0.221 | 2.586 | 0.010 |
| Unfavourable Working Conditions -> COVID-19 Confirmed Cases | 0.211 | 2.695 | 0.007 |
| No. | Missing | Mean | Median | Min | Max | ST | Excess Kurtosis | Skewness | |
|---|---|---|---|---|---|---|---|---|---|
| CC_1 | 1 | 0 | 2338.201 | 2411.11 | 582.2 | 4489.58 | 887.987 | −0.711 | 0.04 |
| CC_2 | 2 | 0 | 19.103 | 20.15 | 7.15 | 29.01 | 5.921 | −0.907 | −0.459 |
| CC_3 | 3 | 0 | 1981.52 | 1928.558 | 220.588 | 6084.265 | 966.515 | 2.177 | 1.059 |
| DR_1 | 4 | 0 | 99.175 | 78 | 0 | 407 | 79.499 | 1.001 | 1.107 |
| DR_2 | 5 | 0 | 193.888 | 182.63 | 0 | 660.58 | 104.839 | 2.097 | 1.034 |
| PLC_1 | 6 | 0 | 13694.64 | 11480.91 | 1977.401 | 112906.2 | 10593.03 | 44.663 | 5.326 |
| PLC_2 | 7 | 0 | 9973.802 | 5781.961 | 0 | 131408.2 | 13802.07 | 34.088 | 4.649 |
| PLC_3 | 8 | 0 | 117452.7 | 80203.89 | 494.983 | 1427001 | 160010 | 29.536 | 4.572 |
| PLC_4 | 9 | 0 | 97409.7 | 73698.64 | 8511.029 | 950871.9 | 104465.2 | 28.552 | 4.428 |
| PLC_5 | 10 | 0 | 11.1 | 11.105 | 9.959 | 12.429 | 0.463 | 0.073 | 0.012 |
| UHC_1 | 11 | 0 | 433.02 | 417.4 | 115.844 | 1335.971 | 152.892 | 10.216 | 2.247 |
| UHC_2 | 12 | 0 | 773.823 | 743.746 | 129.676 | 2399.253 | 273.918 | 9.845 | 2.065 |
| UHC_3 | 13 | 0 | 1523.201 | 1492.594 | 259.351 | 3774.965 | 567.663 | 2.778 | 1.019 |
| UHC_4 | 14 | 0 | 4054.534 | 3817.135 | 1193.017 | 12654.65 | 1435.892 | 11.097 | 2.417 |
| UHC_5 | 15 | 0 | 3487.965 | 3349.508 | 1400.498 | 8241.77 | 1250.983 | 1.898 | 1.159 |
| UHC_6 | 16 | 0 | 785.121 | 753.325 | 168.578 | 2066.574 | 267.522 | 4.379 | 1.275 |
| UWC_1 | 17 | 0 | 43693.91 | 42061.98 | 6160.973 | 169156.3 | 17388.53 | 15.321 | 2.612 |
| UWC_2 | 18 | 0 | 42734.38 | 33224.94 | 870.056 | 170419.6 | 34354.34 | 1.526 | 1.255 |
| UWC_3 | 19 | 0 | 13479.13 | 10843.66 | 1533.984 | 43606.18 | 8429.226 | 1.729 | 1.362 |
| UWC_4 | 20 | 0 | 273931.4 | 266701.6 | 23491.53 | 755759.6 | 133521 | 1.334 | 0.82 |
| Mean | Median | Min | Max | Standard Deviation | Excess Kurtosis | Skewness | Number of Observations Used | |
|---|---|---|---|---|---|---|---|---|
| COVID-19 Confirmed Cases (CC) | 0 | 0.092 | −2.101 | 2.817 | 1 | −0.361 | −0.127 | 177 |
| Death Rate (DR) | 0 | −0.132 | −1.736 | 3.047 | 1 | 0.498 | 0.776 | 177 |
| Living Conditions (PLC) | 0 | −0.325 | −1.034 | 6.816 | 1 | 17.68 | 3.525 | 177 |
| Underlying Health Conditions (UHC) | 0 | −0.084 | −2.182 | 4.95 | 1 | 5.474 | 1.524 | 177 |
| Working Conditions (UWC) | 0 | −0.047 | −1.645 | 3.305 | 1 | 1.218 | 0.976 | 177 |
Acknowledgements
The authors would like to sincerely thank all the reviewers of this paper for their constructive feedback. It is highly appreciated.
Disclosure statement
No potential conflict of interest was reported by the author(s).
References
- Agency for Toxic Substances and Disease Registry. (2016). The social vulnerability index (SVI): data and Tools Download. CDC. [Google Scholar]
- Alegria, M. (2020). Language barriers can mean life or death in fight against coronavirus. https://thehill.com/changing-america/opinion/494284-isolated-with-no-family-members-for-non-english-speakers-with. [Google Scholar]
- Amin, F. (2020). Low-income neighbourhoods at higher risk of coronavirus - CityNews Toronto. [Google Scholar]
- Appropriations, H. C. o. (2020). House Passes Families First Coronavirus Response Act. https://appropriations.house.gov/news/press-releases/house-passes-families-first-coronavirus-response-act. [Google Scholar]
- Bartel, A., Kim, S., Nam, J., Rossin-Slater, M., Ruhm, C., & Waldfogel, J. (2019). Racial and ethnic disparities in access to and use of paid family and medical leave: Evidence from four nationally representative datasets. Monthly Labor Review, https://doi.org/10.21916/mlr.2019.2 [Crossref], [Web of Science ®], [Google Scholar]
- Bartsch, S., Weber, E., Büttgen, M., & Huber, A. (2020). Leadership matters in crisis-induced digital transformation: How to lead service employees effectively during the COVID-19 pandemic. Journal of Service Management, https://doi.org/10.1108/JOSM-05-2020-0160 [Crossref], [Web of Science ®], [Google Scholar]
- Bentler, P. M., & Bonett, D. G. (1980). Significance tests and Goodness of Fit in the analysis of covariance structures. Psychological Bulletin, 88(3), 588–606. https://doi.org/10.1037/0033-2909.88.3.588 [Crossref], [Web of Science ®], [Google Scholar]
- Bhattarai, A. (2020). As more grocery workers die, many supermarket employees fear showing up during pandemic. The Washington Post. [Google Scholar]
- Bolin, B., & Kurtz, L. C. (2018). Handbooks of Sociology and social research. Race, Class, Ethnicity, and Disaster Vulnerability, 181–203. https://doi.org/10.1007/978-3-319-63254-4_10 [Google Scholar]
- Bor, J., Cohen, G. H., & Galea, S. (2017). Population health in an era of rising income inequality: USA, 1980–2015. The Lancet, 389, 1475–1490. https://doi.org/10.1016/S0140-6736(17)30571-8 [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Bravo, M. A., Batch, B. C., & Miranda, M. L. (2019). Residential racial isolation and spatial patterning of hypertension in Durham, North Carolina. Preventing Chronic Disease, 16(3), https://doi.org/10.5888/pcd16.180445 [PubMed], [Google Scholar]
- CDC. (2017). 500 cities project local data for better health. Centers for Disease Control and Prevention. [Google Scholar]
- CDC. (2020a). Cases in the U.S. CDC. Center for Disease Control. [Google Scholar]
- CDC. (2020b). COVID-19 in Racial and Ethnic Minority Groups. https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/racial-ethnic-minorities.html. [Google Scholar]
- CDC. (2020c). COVID-19 pandemic: Lessons learned and future directions. Indian Journal of Ophthalmology, 68, 703–710. https://doi.org/10.4103/ijo.IJO_843_20 [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Centers for Disease Control and Prevention. (2020). COVID-19 in racial and ethnic minority groups CDC. CDC. [Google Scholar]
- Ciotti, M., Ciccozzi, M., Terrinoni, A., Jiang, W.-C., Wang, C.-B., & Bernardini, S. (2020). The COVID-19 pandemic. Critical Reviews in Clinical Laboratory Sciences, 57(6), 365–388. https://doi.org/10.1080/10408363.2020.1783198 [Taylor & Francis Online], [Web of Science ®], [Google Scholar]
- Clark, A., Jit, M., Warren-Gash, C., Guthrie, B., Wang, H. H. X., Mercer, S. W., Sanderson, C., McKee, M., Troeger, C., Ong, K. L., Checchi, F., Perel, P., Joseph, S., Gibbs, H. P., Banerjee, A., & Eggo, R. M. (2020). Global, regional, and national estimates of the population at increased risk of severe COVID-19 due to underlying health conditions in 2020: A modelling study. The Lancet Global Health, 8(8), https://doi.org/10.1016/S2214-109X(20)30264-3 [Crossref], [Web of Science ®], [Google Scholar]
- Clarke, H., & Whiteley, P. (2020). Economic inequality can help predict COVID-19 deaths in the US. USAPP. [Google Scholar]
- Cohen, J. (1988). Statiscal power analysis for the behavioral sciences. Journal of Petrology, 369(1), 1–30. https://doi.org/10.1017/CBO9781107415324.004 [Google Scholar]
- Courts, C. (2020). Language access. California Courts the Judicial Branch of California. https://www.courts.ca.gov/languageaccess.htm. [Google Scholar]
- Cropley, J. (2018). Study finds N.Y. Income disparity greatest in nation. The daily Gazette. The Daily Gazette. [Google Scholar]
- Cunningham, T. J., Croft, J. B., Liu, Y., Lu, H., Eke, P. I., & Giles, W. H. (2017). Vital signs: Racial disparities in age-specific mortality among blacks or African Americans — United States, 1999–2015. MMWR. Morbidity and Mortality Weekly Report, 66, 444–456. https://doi.org/10.15585/mmwr.mm6617e1 [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Edomt. (2020). Covid-19 data. Github. https://github.com/owid/covid-19-data/blob/master/public/data/testing/covid-testing-latest-data-source-details.csv. [Google Scholar]
- Gbongli, K., Xu, Y., & Amedjonekou, K. M. (2019). Extended technology acceptance model to predict mobile-based money acceptance and sustainability: A multi-Analytical structural equation modeling and neural network approach. Sustainability, 11(13), 3639. https://doi.org/10.3390/su11133639 [Crossref], [Web of Science ®], [Google Scholar]
- Green-Laughlin, D. (2020). COVID-19: A Closer Lens. Issues in Mental Health Nursing, 41(8), 662–664. https://doi.org/10.1080/01612840.2020.1773736 [Taylor & Francis Online], [Web of Science ®], [Google Scholar]
- Hair, J. F., Risher, J. J., Sarstedt, M., & Ringle, C. M. (2019). When to use and how to report the results of PLS-SEM. European Business Review, 31(1), 2–24. https://doi.org/10.1108/EBR-11-2018-0203 [Crossref], [Web of Science ®], [Google Scholar]
- Hawkins, R. B., Mehaffey, J. H., Charles, E. J., Schneider, E. B., & Tracci, M. C. (2019). Socioeconomically distressed communities index independently predicts major adverse limb events after infrainguinal bypass in a National Cohort. Journal of Vascular Surgery, 69(1), e12–e13. https://doi.org/10.1016/j.jvs.2018.10.025 [Crossref], [Web of Science ®], [Google Scholar]
- Hearst, M. O., Oakes, J. M., & Johnson, P. J. (2008). The effect of racial residential segregation on black infant mortality. American Journal of Epidemiology, 168(11), 1247–1254. https://doi.org/10.1093/aje/kwn291 [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Henseler, J., Ringle, C. M., & Sinkovics, R. R. (2009). The use of partial least squares path modeling in international marketing. Advances in International Marketing, 20, 277–319. https://doi.org/10.1108/S1474-7979(2009)0000020014 [Crossref], [Web of Science ®], [Google Scholar]
- Heymann, J., Raub, A., Waisath, W., McCormack, M., Weistroffer, R., Moreno, G., Wong, E., & Earle, A. (2020). Protecting health during COVID-19 and beyond: A global examination of paid sick leave design in 193 countries*. Global Public Health, 15(7), 925–934. https://doi.org/10.1080/17441692.2020.1764076 [Taylor & Francis Online], [Web of Science ®], [Google Scholar]
- Ho, R. (2006). Handbook of univariate and multivariate data analysis and interpretation with SPSS. Handbook of Univariate and Multivariate Data Analysis and Interpretation with SPSS, https://doi.org/10.1201/9781420011111 [Google Scholar]
- HRW. (2020). US: Address impact of covid-19 on poor. Human Rights Watch. [Google Scholar]
- Hu, L., & Bentler, P. (1998). Fit indices in covariance structure modeling: Sensitivity to under parameterized model misspecification. Psychological Methods, 3(4), 424–453. https://doi.org/10.1037/1082-989X.3.4.424 [Crossref], [Web of Science ®], [Google Scholar]
- Hussain, S., Fangwei, Z., Siddiqi, A. F., Ali, Z., & Shabbir, M. S. (2018). Structural Equation model for evaluating factors affecting quality of social infrastructure projects. Sustainability, 10(5), 1–25. https://doi.org/10.3390/su10051415 [Crossref], [Web of Science ®], [Google Scholar]
- Jackson, F., & Fitzsimons, V. (2017). Right time, right place. Journal of Christian Nursing : A Quarterly Publication of Nurses Christian Fellowship, 34(4), 259. https://doi.org/10.1097/CNJ.0000000000000429 [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- John, E. (2020). Coronavirus (COVID-19) in the U.S. - Statistics & Facts. https://www.statista.com/topics/6084/coronavirus-covid-19-in-the-us/. [Google Scholar]
- Lohmöller, J.-B. (1989). Latent variable path Modeling with Partial Least Squares. Physica, Heidelberg, https://doi.org/10.1007/978-3-642-52512-4 [Google Scholar]
- Mansoor, S. (2020). Data Suggests Many New York City Neighborhoods Hardest Hit by COVID-19 Are Also Low-Income Areas. Time. https://time.com/5815820/data-new-york-low-income-neighborhoods-coronavirus/ Accessed on April 7, 2020. https://time.com/5815820/data-new-york-low-income-neighborhoods-coronavirus/. [Google Scholar]
- McNicholas, C., & Poydock, M. (2020). Who are essential workers?: A comprehensive look at their wages, demographics, and unionization rates. Economic policy Institute. Economic Policy Institute. [Google Scholar]
- Millett, G. A., Jones, A. T., Benkeser, D., Baral, S., Mercer, L., Beyrer, C., Honermann, B., Lankiewicz, E., Mena, L., Crowley, J. S., Sherwood, J., & Sullivan, P. (2020). Assessing Differential Impacts of COVID-19 on black communities. Annals of Epidemiology, 47, 37–44. https://doi.org/10.1016/j.annepidem.2020.05.003 [Crossref], [PubMed], [Web of Science ®], [Google Scholar]
- Misra, S. (2020). Study finds high rates of positive COVID-19 tests in low-income NYC neighborhoods - ABC News. AbcNews. [Google Scholar]
- National Partnership for Women & Families. (n.d.). Quick Facts. [Google Scholar]
- NYC Health. (2020). Nychealth/coronavirus-data. Department of Health and Mental Hygiene. [Google Scholar]
- Rho, H. J., Brown, H., & Fremstad, S. (2020). A Basic Demographic Profile of Workers in Frontline Industries - Center for Economic and Policy Research. April. [Google Scholar]
- Sammy, H. (n.d.). COVID-19 could close one-third of companies within a month: Survey. Calgary Herald. [Google Scholar]
- Sarstedt, M., & Cheah, J. H. (2019). Partial least squares structural equation modeling using SmartPLS: A software review. Journal of Marketing Analytics, 7(3), 196–202. https://doi.org/10.1057/s41270-019-00058-3 [Crossref], [Google Scholar]
- Statistics, U. B. o. L. (2015). Hispanics and Latinos in industries and occupations. https://www.bls.gov/opub/ted/2015/hispanics-and-latinos-in-industries-and-occupations.htm. [Google Scholar]
- Szabo, L., & Recht, H. (2020). How coronavirus impacts certain races, income brackets, neighborhoods. USA Today. [Google Scholar]
- Taufique, K. (2015). Re: Is it necessary that in model fit my Chi-square value(p-Value) must be non-significant in structure equation modeling (AMOS)? https://www.researchgate.net/post/Is_it_necessary_that_in_model_fit_my_Chi-square_valuep-Value_must_be_non-significant_in_structure_equation_modeling_AMOS/5603fbb86225ffee128b45aa/citation/download. [Google Scholar]
- Tenenhaus, M., Vinzi, V. E., Chatelin, Y. M., & Lauro, C. (2005). PLS path modeling. Computational Statistics and Data Analysis, 48(1), 159–205. https://doi.org/10.1016/j.csda.2004.03.005 [Crossref], [Web of Science ®], [Google Scholar]
- Travis Du Bry. (2014). Fresh fruit, broken bodies: Migrant farmworkers in the United States. Global Public Health, 9(3). https://doi.org/10.1080/17441692.2014.890239 [Web of Science ®], [Google Scholar]
- University of Aberdeen. (2020). Frailty as important as age or underlying health issues in COVID-19 death risk. Medical Xpress. [Google Scholar]
- U.S. Census Bureau. (2018). B19013. Median household income in the past 12 months: 2010 ACS 5-year estimates. Census Reporter. [Google Scholar]
- U.S. Department of Housing and Urban Development. (2014). HUD USPS ZIP Code Crosswalk Files. [Google Scholar]
- Velasquez, D., Uppal, N., & Perez, N. (2020). Equitable Access To Health Information For Non-English Speakers Amidst The Novel Coronavirus Pandemic. https://www.healthaffairs.org/do/10.1377/hblog20200331.77927/full/. [Google Scholar]
- Wetzels, M., Odekerken-Schröder, G., & van Oppen, C. (2009). Using PLS path m modeling for assessing hierarchical construct models: Guidelines and empirical. MIS Quarterly, 33(1), 177–195. https://doi.org/10.2307/20650284 [Crossref], [Web of Science ®], [Google Scholar]
- Wilson, C. (2020). How Coronavirus Hit Low-Income Communities Harder in NYC. Time. [Google Scholar]
- Worldometer. (2020). Coronavirus cases. [Google Scholar]
- Xiao, H., Zhang, Y., Kong, D., Li, S., & Yang, N. (2020). Social Capital and Sleep Quality in Individuals Who self-Isolated for 14 Days during the Coronavirus Disease 2019 (COVID-19) outbreak in January 2020 in China. Medical Science Monitor, 26. https://doi.org/10.12659/MSM.923921 [Crossref], [Web of Science ®], [Google Scholar]
Appendix 1: Indicators descriptive statistics
Table
Appendix 2: Latent variable descriptive
Table