Abstract
In the U.S. health care system, people under age 65 are at risk of losing and regaining health insurance coverage over their lifetimes, which has important consequences for their physical and mental health. Despite the importance of insurance stability, we have an incomplete understanding about the complex factors influencing whether people lose and regain coverage. To advance our understanding of the dynamics of health insurance coverage and guide future research, our purpose is to present a new conceptual model of health insurance stability, where instability is defined as a person’s loss or change of coverage, which can occur more than once in a lifetime. Drawing from theory and evidence in the literature, we posit that personal and plan characteristics, the health system, and the environmental context – economic, social/cultural, political/judicial, and geographic – drive health insurance stability over the life course and are understudied. Studies are needed to identify the populations most at risk of experiencing insurance instability and vulnerability in health outcomes that results from such insecurity, which may suggest reforms and health policies at the individual, health system, or environment levels to reduce those risks.
Keywords
Almost 100 years ago, the U.S. Committee on the Costs of Medical Care recommended health insurance coverage for all Americans. 1 Since that time, universal health insurance coverage has remained an elusive goal for people under age 65 who are ineligible for Medicare. The United States has the lowest population coverage for health insurance among high-income countries. 2 The U.S. has a pluralistic, patchworked system of health insurance, which raises policy concerns about whether Americans have stable coverage they can count on.
Insurance stability occurs when a person has continuous insurance coverage over the course of a lifetime.3,4 The opposite of insurance stability is insurance instability, defined as a person’s loss of coverage, with or without regain of coverage, or a change in coverage; both can occur more than once in a lifetime.3,4 For loss of coverage, the percentage of uninsured U.S. residents has fluctuated greatly over the past two decades. Prior to Congress passing the Patient Protection and Affordable Care Act (ACA) in 2010, about 15% of Americans lacked coverage. 5 In national surveys from 2004–2007, a third of adults under age 65 were uninsured at least 1 month over a 4 year period, and gaps were more frequent in low-income and racial/ethnic minority groups. About 9% lost insurance more than once. For children under age 18, about a quarter lost coverage over a 12 month period.3,4,6 Following passage of the ACA, the uninsured rate declined to 7.9% in 2017 but increased to 8.5% in 2018, the first year-to-year increase since 2008–2009. 5
The percentage of residents who are uninsured fluctuates over time partly because of fluctuations in unemployment rates and the prevalence of employer-sponsored health insurance. 2 About half of the U.S. population is covered by employer-sponsored insurance. 7 Consequently, fluctuations in unemployment rates can produce fluctuations in uninsured rates. For instance, in the 2007–2009 Great Recession, unemployment increased from about 5% in 2008 to almost 10% in 2010, resulting in 9.5 million unemployed workers losing their health insurance.8,9 With the coronavirus (COVID-19) pandemic shutdown of the U.S. economy, which increased unemployment to almost 15%, the uninsured rate in the U.S. population was steady at 8.6% in 2020, because of the continuous coverage provisions in Congress’ Public Health Emergency legislation.7,10 However, excluding adults aged 65 and over with universal Medicare insurance, the uninsured rate was 10.7% for people under age 65, suggesting the economic shutdown impacted uninsured rates in the working population.7,10
Being insured is important because gaps in health insurance coverage are harmful to people’s health.6,11–14 Thornton and Rice 15 estimate that extending private insurance coverage to the uninsured population would save 75,000 lives annually. People benefit most from health insurance when coverage is continuous and long-lasting. 16 Continuous, long-lasting coverage promotes continuity of care, which studies have shown is associated with greater patient satisfaction, better quality of care, fewer emergency department visits and hospitalizations, a reduced risk of mortality, and lower costs of care.17–20 This is the case for children as well. Children with insurance gaps are less likely to have a usual source of care, seek medical care, have preventive visits, and receive vaccinations on time.21–23 People also benefit from the elimination of the psychological stress of being uninsured and reductions in financial risk from health care expenses. 24
Insurance instability can also occur when individuals change health plans. Plan changes often entail shifts in delivery systems or provider networks. As a result, individuals are often forced to move to a different provider, disrupting patient-provider relationships and continuity of care and thereby harming access and quality. 25 For those with private insurance, individuals who changed health plans were twice as likely to change providers, compared to those who did not change plans. 25 Changing health plans may also lead to higher out-of-pocket costs, for example, from starting new deductibles and paying more to stay with providers who are out-of-network in the new plan. 4
Health insurance instability will likely remain an important U.S. policy issue in the foreseeable future. Unfortunately, developing new and innovative strategies to improve stability is challenging in the fragmented U.S. insurance system. We have few insights about the populations and groups most at risk of experiencing insurance instability and vulnerability in health outcomes that results from such insecurity. Better insights may suggest insurance reforms and other interventions at the individual, system, or environment levels to reduce those risks.
To advance our understanding of the dynamics of health insurance coverage and guide future research, our purpose is to present a new conceptual model of health insurance stability. The drivers of health insurance coverage are complex.12,13,26 Drawing from theory and evidence in the literature, we posit that an individual’s personal and plan characteristics are associated with losing or regaining coverage over the life course.13,26 We also posit that environmental context, health policy, and the care delivery system jointly shape individuals’ access to health-promoting resources, including access to jobs and health care facilities that, in turn, influence access to health insurance. The inter-play of these forces may call for inter-related interventions at the individual, health system, or environmental levels to reduce health insurance instability, from birth to death, for everyone in a population.
Conceptual model of health insurance stability
Figure 1 presents our conceptual model of the dynamics of health insurance stability over the life course, which builds on the previous conceptual models in Hill and Schaefer,
26
the Institute of Medicine,12,27 Hadley,
13
Eisenberg and Power,
28
Carrillo and colleagues,
29
Andersen,
30
and Gold.
31
At the center of model is a person’s loss or change of coverage, or insurance instability, at a point-in-time. Our model posits that a person’s likelihood of insurance instability is driven by the static and dynamic personal factors on the left side of the model, the health system factors on the right side of the model, the over-arching environmental context that impinges on all factors in the model, and the interactions between them. With loss or change of coverage, the three factors also jointly influence whether a person subsequently regains coverage and the quality of insurance an individual shifts to or gains. Because insurance instability can occur more than once, over time the losses and changes accumulate and define an individual’s (in)stability of coverage since birth.
27
Conceptual model of the dynamics of health insurance stability over the life course.
Previous conceptual models and studies have focused on the left-side of the model, that is, the personal and plan characteristics affecting coverage loss, while the environmental context has received much less attention.12,13,26,27 The environmental context, along with health policy and the care delivery system on the right side of Figure 1, jointly structure individual and population access to health insurance. The environmental context consists of several inter-related dimensions: economic, social, political – judicial, and geographic. The U.S. political system, in principle, could maximize access to health insurance through legislation and health policy authorizing health plans that cover everyone and can withstand legal challenges, such as the Supreme Court challenges of the ACA. 32 Coverage loss would be rare, and health insurance stability would be high. Instead, in the current, patchworked system of public and private health insurance, access to health plans is restricted through eligibility rules based on employer benefits or a person’s age (Medicare age 65), income (Medicaid low-income threshold), health (Medicaid disability), provider networks, affordability, and other factors. People enjoy stable coverage when they continuously satisfy the eligibility rules of their plans; people usually experience unstable coverage when breaks occur in meeting eligibility rules.
Environment and system levels
The environmental context of where a person lives may be associated with losing or regaining insurance. Local economic characteristics may influence insurance stability. Unemployment rates at the county level can influence the likelihood of securing employer-sponsored coverage. 26
The social environment may also be associated with insurance stability. Neighborhood socioeconomic disadvantage could be associated with insurance instability because disadvantaged areas also tend to have less access to stable job opportunities,33,34 reducing access to employer-provided health insurance. Disadvantaged areas also have limited access to the health care and community resources that residents could leverage to avoid coverage loss or regain insurance.35,36
Another dimension of the social environment is social capital, which may also be associated with insurance stability. County-level social capital refers to the social connections among county residents that can provide material, affective and informational resources to address individual needs.37–39 Places with well-established social networks may offer residents access to health-promoting activities and services, which may reduce their risks of losing coverage or, if uninsured, increase the likelihood of regaining coverage. Areas with high social capital could also facilitate employment, given that social networks have been found to be important resources for individuals hearing about, applying to, and being offered employment.40,41
Similarly, residential racial/ethnic isolation, or the percentage of individuals from racial/ethnic minority groups in a neighborhood, is a cause of racial/ethnic differences in socioeconomic status because predominantly minority areas are more likely to experience business and government disinvestment and receive less educational funding, which may also limit access to employer-sponsored health insurance.42,43
Geographic factors may be associated with insurance stability because access to private insurance is greater in urban than rural areas. In 2010–2019 the uninsured rates were 2–3 percentage points higher in rural than urban areas. 44 Health policy may change this pattern: under the American Rescue Plan, 65% of the 1.9 million rural uninsured individuals who live in HealthCare.gov states have access to at least one zero-premium plan in the insurance exchange. 44
Health policy may influence access to health insurance and insurance stability. Federal, state and local governments are the largest payers and major providers of health care through Medicare, Medicaid, health insurance for public employees, the Veterans Health Administration, the Indian Health Service, state-sponsored health plans, local health departments, community health centers, and other public mechanisms. Governments have an important role in shaping health policy to improve access to health insurance through their legislative and regulatory authority and substantial financial clout: in 2019 governments accounted for almost half (45%) of the $3.8 trillion in U.S. health care spending. 45
Historically, governments have improved access to health insurance through incremental policies rather than universal policies covering everyone.46,47 For example, most states took advantage of the Affordable Care Act to extend Medicaid coverage to adults below 138% of the federal poverty level. Recent findings indicate that Medicaid expansion has decreased coverage disruption in low-income adults.48,49 However, 12 states did not extend Medicaid coverage. If those states extend Medicaid in the future, 2.2 million uninsured adults aged 19 to 64 would gain coverage. 50
Another example is health policy that buffers the loss of employer-sponsored coverage. In 2020 the COVID-19 pandemic shutdown of the economy increased unemployment to almost 15% and threatened widespread loss of employer-sponsored health insurance.7,10,51 In response, Congress passed the American Rescue Plan, which offered premium tax credits to subsidize purchasing coverage from insurance exchanges, mainly for adults under age 65 who are income-ineligible for Medicaid. 47 Congress continues to debate various forms of universal health insurance, such as Medicare for All. 52 In the meantime, Colorado, Nevada and Washington are expanding coverage through their respective state-run health plans. 53 By reducing the percentage of U.S. residents with unstable coverage through these and other policies, the personal characteristics in the conceptual model would become less important as coverage becomes more equitable in the U.S. population.54,55
The care delivery system may influence access to health insurance. When continuous coverage is a goal of the delivery system, health plans and health care organizations have more incentive to invest in their members’ long-term health. For example, when coverage is long-lasting, payers may be more likely to invest in preventive services and reap the health benefits of those investments. 56 From a system perspective, payers benefit from long-term continuous coverage through lower administrative costs of enrollment and disenrollment. 56 If people who lose and regain coverage return to the same source of commercial coverage or delivery system, commercial insurers and systems may have more incentives to invest in preventive services that have long-term benefits. 57 In commercial plans, 12–14% of members in individual and group plans who left their insurer reenrolled after 1 year, 32–34% reenrolled after 5 years, and 47% reenrolled after 10 years. 58
The mix of product types in local markets may influence coverage stability. 59 As managed care penetration increases in areas, market competition for enrollees may increase, with coverage instability as a byproduct, although the effect appears to be small. 25 In areas with high penetration of health maintenance organization (HMO) and point-of-service (POS) plans, people are more likely to switch to HMO plans, which have lower costs, while low-penetration areas have higher disenrollment from HMO plans. 25
Similarly, health insurance exchanges and market competition among managed care plans may contribute to insurance instability. In the ACA Medicaid expansion and the federal and state health insurance exchanges, health plans compete for enrollees under age 65 who are ineligible for Medicaid and not covered by employer-sponsored insurance. 60 On one hand, market competition is a key approach for improving value through lower prices and higher quality. 61 On the other hand, market competition and the exchanges may increase plan switching and instability among private plans in exchanges, including among those private managed care plans competing to enroll Medicaid beneficiaries, and can contribute to individuals churning between Medicaid and private exchange plans due to different eligibility criteria, cost-sharing and premiums, and other factors. 57
Geographic areas with low access to health insurance can be offset by ‘safety net’ community health clinics and public health hospitals offering free or low-cost care to people without insurance. The Veterans Health Administration plays the same role for eligible veterans.
Person level
Returning to the person (left) side of Figure 1,
Income, which is determined partly by education and occupation, may influence an individual’s point-in-time plan characteristics through income-related eligibility requirements. Income changes over time cause “churning,” or insurance instability from transitions in-and-out of Medicaid or between Medicaid and private insurance offered through insurance exchanges or an employer.63,64 For gender, women with private insurance were less likely to lose coverage than men, but women with public coverage were more likely to lose and regain coverage. 6 Race/ethnicity could also play a role; coverage varies by racial/ethnic group, with Whites most likely to have coverage (95%) and have private coverage (75%) and Hispanics least likely (82% and 50%), respectively. 65 Between 2017 and 2018, coverage for Hispanics declined by 1.6 percentage points, but the changes for other groups were not significant.
Immigrant status also defines social position and is associated with health coverage, which is much lower in the foreign-born population (81%) than the native-born population (93%). 65 Areas with greater immigrant populations may be associated with less coverage because of different cultural norms of health care, less knowledge of the U.S. health system, and greater language and economic barriers.13,66 In the ACA era, undocumented immigrants are less likely to be covered partly because they are ineligible for public insurance and plans offered through ACA exchanges. 67 About 42% of undocumented, nonelderly immigrants are uninsured. 68
For working adults, employment may influence an individual’s point-in-time plan characteristics through whether the employer offers insurance, which is more likely with larger employers. 27 In some cases, however, families decline offers of private insurance, due to the high cost of purchasing employment-related insurance and weak preferences for coverage, relying instead on local community clinics for care. 70 For workers with employer-sponsored insurance, job loss causes coverage loss.7–10,27 Ample opportunity exists for the loss and regain of employer-sponsored insurance because U.S. workers change jobs frequently, not all employers offer insurance, and the prevalence of contract, seasonal and gig work jobs, which are less likely to offer insurance, has increased dramatically in recent years and may contribute to Medicaid and insurance exchange churning. 71 In a national longitudinal survey, individuals who were born in 1957–1964 were followed for 39 years, revealing that adults held an average of 12.4 jobs from ages 18 to 54. 72 Changes in employment create risks for losing or changing employer-sponsored insurance, and these risks accumulate over the course of a lifetime.
A person’s family position (employee, spouse, child) and size determine whether a private health plan offers single or family coverage and may shape the risks of losing/gaining insurance if a change in marital status, such as divorce, occurs, because some adults obtain coverage through their spouse.26,65
Health status may play a role: less healthy individuals are more likely to lose insurance, but the association is complex.6,12 One explanation for this association is that the relationship may be spurious—social position is strongly related to both health and insurance coverage as indicated in Figure 1.12,27 However, in their national study, Cutler and Gelber 6 report that adults with worse health were more likely to lose coverage, controlling statistically for educational attainment, indicating that social position may not entirely explain the association. A second possibility is that two-way causation may exist between health status and insurance status: people are unhealthy because they lost insurance.12,27 Third, people purchasing individual health insurance may be denied coverage due to pre-existing conditions. 27 However, because the ACA prohibits insurers from denying coverage to people with pre-existing conditions, this explanation may be inapplicable in the post-ACA era. Fourth, people with low incomes and worse health may be uninsured if they cannot afford the premiums of individually purchased health insurance, or if they do not satisfy Medicaid disability or income eligibility requirements. 12 Countervailing forces also complicate the health status – insurance status relationship. For example, people with a disability are more likely to have public than private coverage (due to Medicaid eligibility for disability), which may improve, rather than reduce, stability. 65 Few studies have examined these relationships in the post-ACA era, and therefore, it is unclear the extent to which health status and insurance status have influenced one another in recent years.
Health status is also related to health care utilization and costs: less healthy individuals have greater utilization and costs. 30 People with more chronic conditions may be more motivated to retain coverage to assure access to health care and protect their economic security.12,27 Few studies have examined whether past utilization or costs predict future loss of coverage, which may be confounded by health status, social position, or other factors in the conceptual model (Figure 1). In contrast, the reverse relationship is well-studied: breaks in coverage are related strongly to reduced utilization. 12 The relationship includes preventive services: in a national sample of adults in the U.S. Health and Retirement Survey, intermittent lack of insurance coverage was associated with less use of preventive services, which can have negative health consequences. 73
Medicaid has a “built-in source of instability” as people’s incomes shift above and below Medicaid’s income eligibility threshold, 75 although individuals in Medicaid expansion states may have less discontinuous coverage than individuals in non-expansion states. 48 Ideally, individuals who are above the income threshold for Medicaid can qualify for subsidized exchange plans, which can reduce breaks in coverage. However, a low-income individual may be unable to purchase an exchange plan because those plans have higher cost-sharing than Medicaid, which can lead to a break in coverage. Another factor is that the Medicaid provider network may not overlap with the provider networks in the exchange plans, which undermines continuity of care.64,76–78
Product type (such as health maintenance organization, point-of-service plan, preferred provider organization, and fee-for-service plan) may be associated with insurance stability. For instance, group and staff model HMOs typically offer generous coverage of preventive services, with the goal of retaining enrollees for many years, which would increase the payoff of their investments in prevention and possibly attract healthier, more preventive-oriented enrollees. 59
Almost all employer-sponsored plans, Medicare, and exchange plans have premiums and cost-sharing (copayments, coinsurance, and deductibles). 79 Prior to the coronavirus pandemic, 22 of the 39 Medicaid-expansion states have obtained federal waivers to impose premiums or cost-sharing for selected services and populations. 80 Premiums and cost-sharing are intended to control health care utilization and expenditures but can have unintended consequences for insurance stability in low-income populations. 81 For adults and children with low-incomes and enrolled in Medicaid, studies consistently find that premiums and cost-sharing are associated with increased financial burden and disenrollment.82,83
Individuals with dual coverage from multiple plans may be less likely to experience breaks in coverage if an individual losses one plan but retains the second plan. For employer-sponsored insurance, employer size is related to insurance instability: compared to large private and government employers, small employers are more likely to drop coverage, and also are more likely to offer fewer insurance options with less generous coverage and higher premiums that employees cannot afford.84,85
Returning to Figure 1, individuals may experience coverage losses and changes more than once, which accumulate over time and define an individual’s (in)stability of coverage since birth (see bottom of Figure 1). Given evidence that being uninsured is associated with worse health, the framework implies that as insurance instability accumulates over the life course, a person’s health may decline, which is captured by the two-way arrow between insurance stability and health status in the conceptual model.
Discussion
In the U.S. health system, people under age 65 are at risk of losing or changing health insurance coverage over their lifetimes, which has important consequences for their physical and mental health outcomes. Despite the importance of insurance stability, we have an incomplete understanding about the complex factors influencing whether people lose or change coverage over the life course, particularly in the post-ACA era. Few studies have examined in depth whether personal and plan factors, the health system, and the environmental context in the conceptual model are associated with insurance instability. We also know little about whether the factors associated with short breaks in coverage overlap with those associated with long-term uninsured spells. 26 Another research area is whether payers benefit from continuous coverage through lower administrative costs of enrollment and disenrollment.
Past studies have focused on whether personal characteristics and insurer type (private vs public) are related to being uninsured. In contrast, the roles of health plan characteristics and, especially, environmental context for insurance instability are understudied. The environmental context can shape individuals’ access to health-promoting resources, including their access to jobs and health care facilities that, in turn, influence access to health insurance. Furthermore, the role of environmental characteristics for health insurance stability may differ depending on the characteristics and resources of individuals and their plans, similar to how prior studies have found that gender differentially impacts insurance stability depending on plan characteristics. 6 Environmental context may, therefore, be an important and underappreciated factor shaping insurance stability.
Lack of health insurance data is a major barrier to studying insurance instability over time on a population level. The U.S. lacks a national information system that tracks the losses and changes of employer, public, or individual-purchased insurance on an individual level from birth through age 65, after which point Medicare offers continuous coverage to eligible residents. Consequently, simple estimates do not exist on the percentage of U.S. residents with continuous versus non-continuous coverage since birth, or the number of coverage gaps or switches in a lifetime.
Barriers also exist in collecting indicators of personal characteristics, such as income or employment status, in the U.S. health care system. National standards do not exist for defining and collecting indicators about personal characteristics in health insurance or medical records.86,87 Another barrier is that indicators of plan characteristics are defined differently and are not available in all health plans. In short, future studies must overcome longitudinal data limitations in the time-varying characteristics and outcomes in the conceptual model. For studies with accurate and complete longitudinal data over a prolonged time period, life course epidemiology offers one approach for examining the consequences of the accumulation of insurance instability over the life course.88,89
In conclusion, the U.S.’s pluralistic, patchworked system of health insurance creates unstable coverage, which has consequences for the health care and health status of the population. Because the fragmented system will likely continue in the foreseeable future, further studies are needed to increase our understanding of the complexity of insurance instability, identifying the populations most at risk of experiencing insurance instability, and launching reforms and policies at the individual, system, or environment levels to close those gaps.
Footnotes
Acknowledgements
The authors thank the anonymous reviewers for their detailed and comprehensive comments about the conceptual model, which have greatly improved the quality of our manuscript.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
