Interaction Between Income, Health Insurance, and Self-rated Health: a Path Analysis

DOI: https://doi.org/10.21203/rs.3.rs-537237/v1

Abstract

Aim: The public health literature suggests health insurance and socioeconomic status (SES) are independent predictors of health outcomes; few studies have explored the interrelationships between these predictors and health. This study explores relationships between income, health insurance, and health, testing the following hypotheses: 1) people with health insurance have higher self-rated health than those who do not; 2) people who have higher income are more likely to have both insurance and higher self-rated health.

Subject and Methods: This is retrospective, cross-sectional, secondary data analysis of 39,450 records from the 2016 National Health Interview Survey (NHIS). The design utilizes path analysis to simultaneously assess relationships between health insurance status, income, and self-rated health, along with several socio-demographic covariates.

Results: We find that higher income and having insurance were both significant predictors of higher self-rated health. Income had a stronger direct effect on health than being insured, as indicated by standardized coefficients. Higher income was also related to having health insurance, thereby influencing health indirectly through its relationship with health insurance. Several socio-demographic variables were also related to self-rated health, income, and insurance.

Conclusions: Future research should explore effects of both insurance and income and their interrelationships on health. Health policies should consider that growing poverty and income inequality in the U.S. limit the effectiveness of health insurance and other social programs if larger social conditions are not addressed.

Introduction

There are few arguments against the notion that access to health insurance has the potential to reduce medical costs to the individual, thereby increasing access to medical services and improving health (Kirby & Kaneda, 2010). Health insurance is specifically designed to subsidize the cost of accessing medical services, thus reducing the financial burden of healthcare to the consumer; this is especially important for the poor and the sick

However, despite the access and financial benefits of health insurance, being insured does not consistently predict better health (Kronick, 2009). One potential reason is that health is influenced by a larger construct: socioeconomic position. Socioeconomic position is known to influence perception of care, health behaviors, and health outcomes (Braveman, Egerter & Williams, 2011).

Prior studies find that higher socioeconomic status and having health insurance are separately associated with better health outcomes, but few examine these relationships as interrelated. Few studies specifically examine whether the positive effect of health insurance on health is connected to socioeconomic status. This study hypothesizes that: 1) people who have insurance have higher self-rated health than those who do not; 2) people who have higher income are more likely to have both insurance and higher self-rated health. Substantial policy implications emerge from these hypotheses, especially given increasing income disparity and ongoing debates about universal health insurance in the United States (U.S.).

Literature Review

In the following subsections we review literature pertinent to our research questions.

Insurance and Health

In the past 20 years, numerous studies have found that being insured is related to better access to healthcare, and having a usual source of care, regular preventive care, and fewer avoidable hospitalizations (Hoffman & Paradise, 2008; Kirby & Kaneda, 2010; Ross & Mirowsky, 2000). Those who are insured have lower mortality and morbidity, higher health-rated quality of life, and increased life expectancy (Arroyave, et al., 2013).

Findings are not unanimous, however. Ross and Mirowsky (2000) find that even with adjustments for initial health, individuals with insurance have worse or no different health than the uninsured. Black et al. (2013) and Kronick (2009) find that controlling for demographic factors, health status, and health behaviors, the uninsured do not have worse health outcomes than the insured. In addition, despite the increase in the percentage of Americans with private health insurance since the 2010 enactment of the Affordable Care Act (ACA), population health has not improved, and life expectancy has decreased slightly (Ho & Hendi, 2018).

The mixed results of studies and the lack of improvement in health suggests that many factors other than health insurance contribute to health, perhaps to an even greater degree.

Income and Health

One factor that is known to be strongly associated with health is socioeconomic status, broadly comprised of elements such as education, employment/occupation, income/wealth and environment, and often represented by income (Cowan, et al, 2012). Studies have found that higher income and other socioeconomic elements are associated with better health status, self-rated health, and life-expectancy (Braveman, Egerter & Williams, 2011; Chetty, et al., 2016; Health Affairs Health Policy Brief, 2018; Pathak, Low, & Swint 2021; Ross & Mirowsky, 2000; Woolf, et al., 2015).

Income and Insurance

Few studies have examined the relationship between income and insurance. Three studies find that income is a predictor of having health insurance (Burtless & Svaton, 2010; Fronstin; 2005; Markowitz, Gold & Rice, 1991).

Pathways between Socioeconomic Status and Health

In explaining why socioeconomic status matters in health, a number of pathways have been considered. Socioeconomic position is thought to influence health through illness/health-promoting resources, experiences, and health behaviors (Braveman, Egerter & Williams, 2011). Concentrated poverty is correlated with economic instability, lack of social solidarity, decreased social capital, and other resources and experiences that contribute to health (Dreier, Mollenkopf, & Swanstrom, 2013). Social policies (poor access to healthcare, food deserts, poor quality housing, etc.) that reinforce the material conditions of concentrated poverty are another partial explanation of the wealth-health phenomenon (Dreier, Mollenkopf, & Swanstrom, 2013).

The Pathway of Income-Insurance-Health

Access to healthcare via health insurance is theorized to be an important pathway between socioeconomic status–particularly income–and health (Braveman Egerter & Williams, 2011). In addition, the inconsistent research findings regarding the relationship between insurance and health may due to the effects of socioeconomic status. For example, for those living in areas of concentrated poverty, quality of care is lower for the insured and the uninsured alike (Dreier, Mollenkopf & Swanstrom, 2013).

To date, the income-insurance-health pathway remains theoretical. The lack of empirical evidence begs the following questions: 1) Does health insurance impact health independently of socioeconomic status, or in relationship with socioeconomic status, or both? 2) Does socioeconomic status (as measured through income) impact health directly, or through its impact on access to health care (health insurance), or both?

Other Factors Impacting Income, Insurance, and Health

This study focuses on the relationships between the socioeconomic element of income, health insurance status and health. Other measures of socioeconomic status, such as education and employment, are interrelated with income and complexly related to having insurance and health status. For example, higher education and employment predict higher income (Hadley, 2003), which predicts having insurance (Markowitz, Gold & Rice, 1991) and better health (Guma, Sole-Auro, & Arpino, 2019). Therefore education and employment are included in our model as influencing all of the three main variables of income, insurance and health.

Other factors that may influence income, health insurance and health are age, ethnicity, race, gender, marital status, geography, and country of birth (U.S. born or not). Age affects the probability of having insurance, since 14.5% of working age adults are uninsured, while all adults 65 and older have Medicare (Cohen, et al, 2020). Older individuals are also more likely higher income, but worse health (Cheng et al, 2013). Blacks and Hispanics tend to have a lower income (Akee, Jones, & Porter, 2019), are less likely to have health insurance (Health Affairs Health Policy Brief, 2018), and are more likely to have poorer health than Whites (Cogburn, 2019). On average, women work for less pay, in smaller firms, with fewer benefits and more part-time work than men (Boniol et al, 2019), so they are more likely to have lower income and less likely to have employer-sponsored health insurance. Despite greater longevity, women have lower self-reported health and other health disparities (United Health Foundation, 2021).

Regarding marital status, married people tend to have higher household incomes (Perry, 2019). Married women are more likely to have employer-sponsored health insurance through either their own employment or that of their spouse (Simpson & Cohen 2017).

Geographical considerations show that mortality is higher in rural compared to urban areas (Gong, et al., 2019). In one study, higher mortality in rural areas is partially explained by socioeconomic deprivation and lack of health insurance (Gong, et al., 2019). Since certain U.S. regions, such as the South, have higher concentrations of rural areas than other regions, these characteristics play out on a regional basis.

Assessing the impact of country of birth is difficult, as income levels, health insurance coverage and health are not uniform across immigrating nationalities. In general, the foreign-born are less likely to have insurance, but the longer they live in the U.S., the more likely they are to obtain it (Greico, 2004). With exceptions for certain nationalities, the foreign-born are also more likely to have lower income (USA Facts, 2019). In one exception, Neilson (2017) reports that foreign-born Blacks have incomes 30% higher than native-born. And while many of the foreign-born may have higher rates of diabetes, infections, and occupational injuries, they tend to have lower rates of mortality, circulatory diseases, overweight/obesity, and some cancers (Argeseanu, Ruben, & Narayan, 2008). Their health, however, resembles natives the longer they are in the U.S. Due to the complexities of country of birth, the direction of the relationship with insurance coverage, income and health is uncertain.

Literature Summary

The literature indicates that researchers have not fully unpacked the relationships between income, health insurance status, and health. Additional studies are necessary to understand whether insurance and income are associated with each other, and together, with health, and to develop a more unified theory of health insurance, socioeconomic status, and health.

Methods

Study design and measures. This is a retrospective, cross-sectional, secondary data analysis of records from the 2016 National Health Interview Survey (NHIS). The design utilizes path analysis to simultaneously assess relationships between health insurance status (insured/uninsured), income (eleven ordinal categories), and self-rated health (five ordinal categories), along with socio-demographic covariates of education, employment, age, race, ethnicity, gender, marital status, region of residence, and U.S. birth. In order to maximize degrees of freedom in the path model, multinomial measures of race, marital status, and employment were re-categorized as dichotomous variables. Region was modeled as one variable having four nominal categories. Variable categories, including reference categories, are noted in Table 2.

Path analyses. At the conclusion of data cleaning, univariate, bivariate, and stochastic imputation analyses, the dataset held 39.450 observations. Descriptive statistics were run, and a multivariate path model was analyzed via SPSS AMOS version 24. Path analysis is the analytical method of choice to test the association between correlated variables in which several endogenous variables are modeled. The initial structural model, presented in Fig. 1, hypothesized the following: all socio-demographic variables predict income, insurance status, and self-rated health; income predicts insurance and self-rated health; insurance predicts self-rated health.

Model Fit and Revisions. Model fit was assessed using several indicators that are available in SPSS AMOS: root mean square error of approximation (RMSEA), with values between 0.05 and 0.10 considered fair fit; comparative fit index (CFI); and three relative fit indices [incremental fit index (IFI), normed fit index (NFI), and the Tucker-Lewis index (TLI)] were all used. Values of the fit indices range from 0.0 to 1.0 and values approaching 1.0 (above 0.9 is best) demonstrate best fit.

The initial model indicated that a few paths were not a good fit for the data. These paths were removed from analysis, and covariance paths were added between several of the covariate variables. To attain the best possible fit, only statistically significant paths were included in the model. The final model had relatively good fit, and all included paths were significant at the 0.01 level. The goodness of fit indices of the final model are in Table 1.

Results

Table 2 provides descriptive statistics. The sample was approximately half male and half female. It was less racially diverse (81% white) than the population (76.6%). Average age was 42.93, more than half were married (57%), almost half (48%) had an associate’s degree or higher, and most (86%) were employed for pay. Most (84 %) were native born. Thirty-three percent described their region of residence to be South, 27% West, 22% Midwest, and 17% Northeast. Nearly half reported an annual income of less than $35,000, while 18% reported income of $75,000 or more. Nearly 90% were insured, slightly less than the 2016 population average (91%). More than two-thirds rated their health as “very good” or “excellent.”

Figure 2 is the final path model and Table 3 provides the significance values for each path relationship. Higher income and having insurance were predictors of higher self-rated health, and higher income contributed to having health insurance. This indicates that income is significantly related to self-rated health both directly and indirectly through insurance. Income also had a stronger direct effect on self-rated health than being insured, as indicated by standardized coefficients.

All socio-demographic variables except gender and region (excluded from final model) were significantly related to self-rated health. Being younger, more educated, white, non-U.S.-born, unemployed, married, and Hispanic were related to higher self-reported health. These findings are intuitive and consistent with the literature.

All socio-demographic variables were statistically related to income. Being employed, having higher levels of formal education, and being male, older, married, non-Hispanic white, from regions other than the South, and U.S.-born were related to having higher income. Employment, education, age, gender, and marital status had higher standardized coefficients, while race, ethnicity, region of residence, and U.S. birth had lower standardized coefficients. These results are consistent with the literature reviewed.

Finally, all socio-demographic variables except race (excluded from final model) were also significantly related to insurance status. Higher education, being U.S.-born, female, married, non-Hispanic, older, and employed were related to having insurance. Education had the highest standardized coefficient, followed by being U.S.-born, female, married, non-Hispanic, older, and employed.

Limitations

The primary constraint of this study was the availability of data. The NHIS dataset is one of the few national datasets that contain health, socioeconomic and socio-demographic status and health insurance records all together. However, health status in the NHIS data was limited to self-rated health. This was a major constraint of this study. Though self-rated health is a significant predictor of objective health status, there may be differences between self-rated and actual health in terms of their relationships to the variables in this study. Data were also not available to consider healthcare utilization and the influence of concentrated poverty on healthcare utilization and self-rated or objective measures of health.

Specific limitations of the measures in the NHIS dataset also limited the analysis. Whether the person lived in an “urban” or “rural” setting was restricted for public use and could not be included in this study. Data on racial concentration or neighborhood level were not available. The dataset did not include a measure of “underinsurance,” which concerns people who have insurance but not enough to cover their basic medical needs. The NHIS also does not report income as a continuous variable and has an open-ended upper category of “$75,000 and above.” This was a major limitation of the dataset considering the main independent variable was earned income.

Discussion

Prior studies find that higher socioeconomic status and having health insurance are separately associated with better health outcomes, but few examine these relationships as interrelated. This study suggests the interrelatedness of health insurance, income and health and the strong influence of income on self-rated health. While this study cannot claim causal mechanisms between health insurance, income, and self-rated health, it adds to the body of knowledge examining whether health insurance has a positive effect on health independent of income, or whether it does so in relationship to income.

As this study indicates that income and health insurance work in tandem to produce self-rated health, future research should explore impacts of both insurance and income on health and the interconnectedness of income, insurance and health, rather than viewing the impact of these factors on health as either/or, or both independently. In addition, as the type or duration of insurance could affect health, studies are also recommended to examine impacts of income on health in conjunction with specific types of insurance and insurance duration.

Health policies should consider that growing poverty and income inequality in the U.S. limit the effectiveness of health insurance and other social programs on health if the larger social conditions are not addressed. Increasing health insurance coverage without addressing the larger socioeconomic context greatly underestimates the limitations of health insurance. Furthermore, failing to ensure access to health insurance while simultaneously not addressing the material conditions of life may have doubly severe consequences for the American people. The strong direct and indirect relationship between income and self-rated health also supports the public health assertion that socioeconomic status is one of the most important determinants of health, regardless of race or ethnicity (Woolf, et al., 2015).

One way to address socioeconomic influences on health is through health impact assessments (HIAs). These assessments enable policy-makers to identify the health effects of proposed new policies. For example, proposed projects that reduce walkable environments or create food deserts would have a negative impact on health and should be discouraged, whereas policies providing income supplementation and educational opportunities for low income families could lead to health benefits for these families and should be encouraged. At this time, most HIAs are performed by community organizations (Pathak, Low, & Swint 2021). State mandates requiring HIAs prior to built environment, energy and transportation projects would promote greater use of these assessments (Pathak, Low, & Swint 2021).

Hospitals can also address socioeconomic determinants of health by including socioeconomic considerations in their Community Health Needs Assessments (CHNAs). Nonprofit hospitals are mandated to conduct CHNAs as a Patient Protection and Affordable Care Act (ACA) mandate for maintaining their tax-exempt status. The assessments should include, as part of the implementation planning process, patient interventions that improve socioeconomic status. Implementation plans that are meaningful, impactful, and demonstrate improved population health outcomes could be a basic requirement to help legislators determine which hospitals are appropriately utilizing the tax relief. Also, as a direct and indirect determinant of health, income should be assessed for inclusion in patient risk scores in population health management systems. As hospital systems develop their population health models, healthcare payers and providers who stratify patient risk for the distribution of clinical resources should consider income as a socioeconomic stratification category.

In order to promote these changes in hospital practice, the ACA may need to be amended to strengthen the tax-exempt hospital requirements. At present, the ACA neither incentivizes tax-exempt hospitals to include socioeconomic considerations in their implementation plans, nor does it penalize hospitals who do not include these considerations. Legislators should consider including additional penalties or incentives in the CHNA section of the ACA for socioeconomic considerations.

Finally, in future healthcare reforms legislators should consider the implications of un- and underinsurance, and insurance without access to providers. At minimum, legislators must incentivize providers to accept Medicaid, require states to expand Medicaid, and explore health equity indices in evaluating Medicaid and Medicare.

Declarations

Funding

This research was conducted as a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy. No funding was procured to conduct this secondary data analysis.

Conflicts of interest/Competing interests

The authors of this manuscript have no conflicts of interest or disclosures.

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and material

The data analyzed in this study are publicly available through the Centers for Disease Control and Prevention’s National Center for Health Statistics. The National Health Interview Survey and corresponding data may be accessed at the following website: https://www.cdc.gov/nchs/nhis/index.htm

Code availability

The authors of this study utilized IBM SPSS Statistics version 24 and IMB SPSS AMOS version 24. Syntax used to conduct this study may be made available to anyone upon request.

Authors' contributions

This is an original study conducted by Atalie Ashley-West in partial fulfilment if the of the requirements for the Doctor of Philosophy degree. She worked with her committee to conceptualize and conduct the study, analyze the data, and draw relevant conclusions. Lynn Unruh was her dissertation committee chair and contributed substantially to the conceptualization of this study and the editing of this manuscript for publication.

References

  1. Akee R, Jones MR, Porter SR. (2019) Race Matters: Income Shares, Income Inequality, and Income Mobility for All U.S. Races. Demography 56(3): 999-1021. https://doi.org/10.1007/s13524-019-00773-7.
  2. Argeseanu CS, Ruben JD, Narayan KM (2008) Health of foreign-born people in the United States: a review. Health Place 14(4): 623-35. https://doi.org/10.1016/j.healthplace.2007.12.002
  3. Arroyave I, Cardona D, Burdorf A, Avendano M (2013) The impact of increasing health insurance coverage on disparities in mortality: health care reform in Colombia, 1998-2007. Am J Public Health 103(3): e100-6.  https://doi.org/10.2105/AJPH.2012.301143
  4. Black B, Espín-Sánchez JA, French E, & Litvak K (2013) The effect of health insurance on near-elderly health and mortality. Northwestern University Law School Law and Economics Research Paper no. 12-09. http://dx.doi.org/10.2139/ssrn.2103669
  5. Boniol M, McIsaac M, Xu L, Wuliji T, Diallo K, Campbell J (2019) Gender equity in the health workforce: analysis of 104 countries. Geneva, CH: World Health Organization https://www.who.int/hrh/resources/gender_equity-health_workforce_analysis/en/.  Accessed 15 Jan 2021.
  6. Braveman P, Egerter S, Williams DR (2011) The social determinants of health: coming of age. Annu Rev Public Health 32: 381–98. https://doi.org/10.1146/annurev-publhealth-031210-101218
  7. Burtless G, & Svaton P (2010) Health care, health insurance, and the distribution of American incomes. Forum Health Econ Policy 13(1): 1-39. https://doi.org/10.2202/1558-9544.1194
  8. Cheng Y, Chen IS, Chen CJ, Burr H, Hasselhorn HM (2013) The influence of age on the distribution of self-rated health, burnout and their associations with psychosocial work conditions. J Psychosom Res 74(3): 213-20. https://doi.org/10.1016/j.jpsychores.2012.12.017
  9. Chetty R, Stepner M, Abraham S, Lin S, Scuderi B, Turner N, Bergeron A, & Cutler D (2016) The Association Between Income and Life Expectancy in the United States, 2001-2014.  JAMA 315(16): 1750–1766. https://doi.org/10.1001/jama.2016.4226
  10. Cohen RA, et al (2020) Health Insurance Coverage: Early Release of Estimates from the National Health Interview Survey, 2019, CDC, National Center for Health Statistics https://www.cdc.gov/nchs/data/nhis/earlyrelease/insur202009-508.pdf. Accessed 15 Feb 2021
  11. Cogburn CD (2019) Culture, Race, and Health: Implications for Racial Inequities and Population Health. Milbank Q 97(3): 736–761. https://doi.org/10.1111/1468-0009.12411
  12. Cowan CD, Hauser RM, Kominski RA, Levin HM, Lucas SR, Morgan SL, …& Chapman C (2012) Improving the measurement of socioeconomic status for the national assessment of educational progress: a theoretical foundation. Recommendations to the National Center for Education Statistics. https://nces.ed.gov/nationsreportcard/pdf/researchcenter/Socioeconomic_Factors.pdf. Accessed 15 Feb 2021.
  13. Dreier P, Mollenkopf J, & Swanstrom T (2013) Place matters: Metropolitics for the 21st century. Lawrence, KS: University Press of Kansas.
  14. Fronstin, P (2005) The relationship between income and health insurance: Rethinking the use of family income in the current population survey. EBRI Notes 26(2): 2-6.
  15. Gong G, Phillips SG, Hudson C, Curti D, Philips BU (2019) Higher US Rural Mortality Rates Linked To Socioeconomic Status, Physician Shortages, And Lack of Health Insurance. Health Aff (Millwood) 38(12): 2003-2010. https://doi.org/ 10.1377/hlthaff.2019.00722
  16. Greico, (2004) Health Insurance Coverage of the Foreign Born in the United States: Numbers and Trends. Migration Policy Institute Fact Sheet. https://www.migrationpolicy.org/research/health-insurance-coverage-foreign-born-united-states-numbers-and-trends. Accessed 28 Feb 2021.
  17. Guma, J, Sole-Auro, A & Arpino, B (2019) Examining social determinants of health: the role of education, household arrangements and country groups by gender. BMC Pub Health. 19: 699.  https://doi.org/10.1186/s12889-019-7054-0
  18. Hadley J (2003) Sicker and poorer--the consequences of being uninsured: a review of the research on the relationship between health insurance, medical care use, health, work, and income. Med Care Res Rev 60(2 Suppl): 3S-75S. https://doi.org/10.1177/1077558703254101
  19. Health Affairs Health Policy Brief (2018) Health, Income & Poverty: Where We Are and What could Help. Health Affairs Health Policy Brief., Oct; https://www.healthaffairs.org/do/10.1377/hpb20180817.901935/full/. Accessed, 15 Feb 2021
  20. Ho, JY, Hendi AS (2018) Recent trends in life expectancy across high income countries: retrospective observational study. BMJ 362: k2562. https://doi.org/10.1136/bmj.k2562
  21. Kirby JB, Kaneda T (2010) Unhealthy and uninsured: exploring racial differences in health and health insurance coverage using a life table approach. Demography 47(4): 1035-51. https://doi.org/10.1007/BF03213738
  22. Kronick R (2009) Health insurance coverage and mortality revisited. Health Serv Res 44(4): 1211-31.  https://doi.org/10.1111/j.1475-6773.2009.00973.x
  23. Markowitz MA, Gold M, Rice T (1991) Determinants of health insurance status among young adults. Med Care 29(1): 6-19. https://doi.org/10.1097/00005650-199101000-00002
  24. Nielsen (2015) Increasingly affluent, educated, and diverse. African-American consumers: The untold story: 2015 report. Nielson Report. http://www.nielsen.com/content/dam/corporate/us/en/reports-downloads/2015-reports/african-american-consumer-untold-story-sept-2015.pdf. Accessed 28 Feb 2021.
  25. Pathak, S, Low, DM & Swint, JM (2021) Trends in the implementation of health impact assessments in the United States: exploring the role of educational attainment, poverty, and government ideology. J Public Health (Berl.) 29, 353–360. https://doi.org/10.1007/s10389-019-01122-z
  26. Perry, M (2019) Explaining US income inequality by household demographics, 2018 update. American Enterprise Institute https://www.aei.org/carpe-diem/explaining-us-income-inequality-by-household-demographics-2018-update/. Accessed 28 Feb 2021.
  27. Ross CE, Mirowsky J (2000) Does medical insurance contribute to socioeconomic differentials in health? Milbank Q 78(2): 291-321, 151-2. https://doi.org/10.1111/1468-0009.00171
  28. Simpson JL, Cohen RA (2017) The Association of Marital Status and Offers of Employer-based Health Insurance for Employed Women Aged 27-64: United States, 2014-2015. NCHS Data Brief (268):1-8.
  29. Woolf, S, Aron, L, Dubay, L,  Simon, S, Zimmerman, E and  Luk, K (2015) How are income and wealth linked to health and longevity? Urban Institute http://webarchive.urban.org/publications/2000178.html. Accessed 28 Feb 2021.
  30. United Health Foundation (2021) Health Disparities by Gender: 2019 Senior Report. United Health Foundation https://www.americashealthrankings.org/learn/reports/2019-senior-report/findings-health-disparities-by-gender. Accessed 28 Feb 2021.
  31. USA Facts. (2019) USA Facts 2019 Annual Report. USA Facts Institute https://annualreport.usafacts.org/articles/52-download-2019-annual-report. Accessed 20 Feb 2021.

Tables

Table 1.

Goodness of Fit Indices of Final Model

Test

Abbreviation

Value

Root mean square error of approximation

RMSEA

0.079

Comparative fit index

CFI

0.831

Normed fit index

NFI

0.831

Tucker-Lewis index

TLI

0.652

Incremental fit index

IFI

0.831

 

Table 2.

Descriptive Statistics, Post Imputation (N = 39,450) 

Variable

Frequency

Percent

Percent National Average*

Gender

 

 

 

 

Male

20,172

51.1

49.2

 

Female

19,279

48.9

50.8

 

Race

 

 

 

 

White

32,126

81.4

76.6

 

Non-White

7,316

18.5

23.2

 

Hispanic Ethnicity

 

 

 

 

Yes

5,588

14.2

18.1

 

No

33,862

85.8

81.9

 

Marital Status

 

 

 

 

Married

22,530

57.1

 

 

Not married

16,920

42.9

 

 

Education

 

 

 

 

Never attended school

87

0.2

 

 

1st-12th grade, no diploma

3,298

8.4

 

 

HS diploma/GED

9,262

23.5

 

 

Some college

7,796

19.8

 

 

Associate’s degree

4,993

12.7

 

 

Bachelor’s degree

8,823

22.4

30.3

(BA or higher)

Master’s degree

3,815

9.7

 

 

Doctorate

1,370

3.5

 

 

Employment Status

 

 

 

 

Employed, for pay

33,812

85.7

63.1

 

Not employed

5,638

14.3

 

 

Born in the U.S.

 

 

 

 

Yes

33,221

84.2

86.8

 

No

6,229

15.8

13.2

 

Geographical Region

 

 

 

 

Northeast

6,796

17.2

 

 

Midwest

8,827

22.4

 

 

South

13,042

33.1

 

 

West

10,785

27.3

 

 

Income

 

 

 

 

$0-$4,999

3,935

10.0

 

 

$5,000-$9,999

2,257

5.7

 

 

$10,000-$14,999

2,601

6.6

11.2

(under $15k)

$15,000-$19,999

2,546

6.5

 

 

$20,000-$24,999

2,843

7.2

9.6

($15-24.9k)

$25,000-$34,999

5,133

13.0

9.4

 

$35,000-$44,999

4,505

11.4

12.9

($35-49.9k)

$45,000-$54,999

3,820

9.7

 

 

$55,000-$64,999

2,749

7.0

 

 

$65,000-$74,999

2,056

5.2

17.0

($50-74.9k)

$75,000 and over

7,005

17.8

40

($75k & over)

Health Insurance Status

 

 

 

 

Uninsured

4,223

10.7

9

 

Insured

35,227

89.3

91

 

Self-Rated Health

 

 

 

 

Poor

271

0.7

 

 

Fair

2,254

5.7

 

 

Good

9,858

25.0

 

 

 

 

Very Good

14,630

37.1

 

Excellent

12,436

31.5

 

 

Minimum

Maximum

Mean

 

Age

18

85

42.93

 

Data Sources:

* 2016 Current Population Survey, U.S. Census Bureau

2016 National Health Interview Survey, Centers for Disease Control and Prevention

Table 3. Path Statistics