Objectives. I examined how labor market and health insurance outcomes were affected by the loss of dependent coverage eligibility under the Patient Protection and Affordable Care Act (ACA).

Methods. I used National Health Interview Survey (NHIS) data and regression discontinuity models to measure the percentage-point change in labor market and health insurance outcomes at age 26 years. My sample was restricted to unmarried individuals aged 24 to 28 years and to a period of time before the ACA’s individual mandate (2011–2013). I ran models separately for men and women to determine if there were differences based on gender.

Results. Aging out of this provision increased employment among men, employer-sponsored health insurance offers for women, and reports that health insurance coverage was worse than it was 1 year previously (overall and for young women). Uninsured rates did not increase at age 26 years, but there was an increase in the purchase of non–group health coverage, indicating interest in remaining insured after age 26 years.

Conclusions. Many young adults will turn to state and federal health insurance marketplaces for information about health coverage. Because young adults (aged 18–29 years) regularly use social media sites, these sites could be used to advertise insurance to individuals reaching their 26th birthdays.

In September 2010, one of the first provisions of the Patient Protection and Affordable Care Act (ACA; Pub L No. 111–148) went into effect, allowing young adults up to age 26 years to remain on a parent’s health insurance plan as a dependent, provided that they did not have an offer for health coverage through their own employer. The goal was straightforward—to expand health insurance to a group of individuals that historically had high rates of being noninsured. Because the predominant source of health insurance in the United States for working-age adults is through an employer,1 this provision relaxed the tie between employment and insurance for young adults and allowed more flexibility in job choice and a potential reduction in job lock, or the inability to leave a job for fear of losing health insurance benefits. For qualifying individuals seeking health insurance, the provision altered the insurance choice set, leading to changes in labor market and health insurance outcomes. Because eligibility for this program expires on an individual’s 26th birthday, these changes are most prevalent at approximately age 26 years. I estimated the impact of turning 26 years, or “aging out,” on labor and health insurance market outcomes for young adults in the United States.

The provision has expanded health coverage to millions of young adults. In 2012, nearly 8 million adults between the ages of 19 and 25 years were able to remain on their parents’ plans.2 Previous work focused primarily on the resultant gains in health insurance coverage for this group as a whole (aged 19–25 years). Using different data sources, all of these studies found gains in health coverage of approximately 3 to 6 percentage points, which showed unequivocally that the provision succeeded at expanding health insurance for the targeted group.3–6 In terms of changes in labor market outcomes, studies found no evidence of the provision changing the likelihood of a young adult being employed, but small reductions were seen in the probability of working full time and the number of hours worked per week.7

Rather than comparing changes in coverage and employment for the entire targeted group (young adults aged 19–25 years) with changes in coverage for older adults (adults aged 26–30 years), I focused on what happened to the young adults at or above the eligibility threshold and to those who aged out of the provision. The natural threshold that occurs at age 26 years as a result of this provision led to variations in the characteristics of the marginally ineligible young adult, providing insights into the labor and insurance market choices that might result when the ACA is fully implemented.

The National Health Interview Survey (NHIS) provides detailed information on health, health insurance, and employment for a representative sample of the overall civilian, noninstitutionalized population of the United States. Data were derived from a harmonized version of the NHIS, the Integrated Health Interview Series (IHIS), provided by the Minnesota Population Center.8

My sample was restricted to the time period after implementation of the provision, but before the ACA individual mandate and expansion of the dependent coverage provision9 (2011–2013). Beginning in 2014, if a young adult was offered coverage through an employer, they were still eligible for coverage through a parent’s plan. Within the NHIS, labor and health insurance outcomes are asked of all individuals, with the exception of whether health coverage type was better, worse, or the same as the previous year, which is limited to a randomly selected sample adult within the household.

I only included nonmarried individuals in the analytical sample for several reasons. Married men have higher labor market participation rates than their unmarried counterparts, whereas married women are less likely to be in the labor market than unmarried women.10 Marital status is also associated with increases in health coverage and employer-sponsored insurance offer rates for women.11 Also, because ages 26.6 and 29.0 years are the average ages of women and men, respectively, at their first marriage (in 2013),12 it was plausible that including married individuals in the sample would lead to a disproportionate amount of unmarried individuals younger than 26 years. As part of robustness checks, I estimated models that included married individuals, and the results were similar to those of the main findings. Previous work revealed different effects based on gender,3,6 so I produced estimates for the full sample and by gender.

An attractive feature of the NHIS is that it contains respondents’ month of birth and interview for each survey year. These can be used to create a more precise definition of age at the time of the interview and to adjust for distance from the adult provision eligibility cutoff. Following studies with a similar methodology,13,14 the sample included respondents who were up to 2 years younger or older than the eligibility threshold of 26 years. The series of questions regarding employment had a 2-week reference period, whereas the health insurance questions referred to the insurance status at the time of survey. Because my empirical strategy compared young adults who were slightly younger than the young adult provision age cutoff with those who were slightly older, these short reference periods for outcome measures were ideal. My final analytical sample included 10 463 unmarried individuals. I weighted the analyses using the survey estimation procedure (SVY) in Stata version 12 (StataCorp, College Station, TX).

My outcomes focused on changes in employment, employment-related health coverage, coverage type, and plan satisfaction. My 3 labor market measures were employed, in the labor force, and employed full-time. I used employer-sponsored insurance (ESI) and an offer of ESI to gauge employment-related health coverage. I captured health coverage type by public, private, or no insurance, and I also measured type of private insurance by nongroup, directly purchased coverage. Lastly, I analyzed an indicator of health plan quality (added to the NHIS in 2011 for sample adults only).

I used a regression discontinuity design to estimate how aging out of the dependent coverage provision affected labor and health insurance outcomes among young adults. This methodology took advantage of the sudden change in health coverage options that might result after an individual turned 26 years old and became ineligible for health insurance through a parent. I followed the methodology of previous regression discontinuity literature (i.e., visual data inspection)13,14 to ensure the smooth profile of age was correctly specified. A quadratic polynomial was the best fit for the data. The model with the age profile fully interacted with the treatment was:

is a labor market or health insurance outcome for individual i. Vector contains observable characteristics for individual i, including dummy variables to control for poverty status. This indicator of poverty status reflects the US Department of Health and Human Services poverty guidelines and was created by the IHIS staff using family size and imputed family income. I opted to use income at or below 138% of the federal poverty guidelines because the public data file did not contain continuous family income and it was important to control for the possibility that poverty influenced both labor market and health insurance outcomes. Thus, family income at or below 138% of the federal poverty guidelines was the best available proxy for poverty. The selection of 138% of the federal poverty guidelines by IHIS staff reflected the guidelines commonly used in administrative purposes for pubic eligibility (in particular, the Medicaid expansion, which allows single, childless adults with incomes at or below 133% of the federal poverty guideline, plus a 5% income disregard to qualify for Medicaid if they reside in a state that agreed to expand the program). Other dummy variables to control for poverty were highest educational attainment, region of residence, health status, presence of a chronic health condition, citizenship, gender (for the models including both men and women), and race/ethnicity. Because my study design took place over 3 years, I included year-fixed effects ().

The treatment measure is captured by , which is zero for all individuals younger than 26 years and 1 for those aged 26 years and older. For all models, agei is the difference in age before or after the individual’s 26th birthday.

I used logit models to estimate and controlled for the complex design of the sample survey by using the survey estimation procedure (SVY) in Stata version 12. The results (Table 1) show the average marginal effect, or percentage point increase (or decrease) in an outcome in response to turning 26 years old. I also performed several robustness checks for model fit and sample design (results can be found as data available as a supplement to the online version of this article at http://www.ajph.org).

Table

TABLE 1— The Effect of Turning 26 Years Old on Labor Market and Health Coverage Outcomes for Unmarried Young Adults: National Health Interview Survey, United States, 2011–2013

TABLE 1— The Effect of Turning 26 Years Old on Labor Market and Health Coverage Outcomes for Unmarried Young Adults: National Health Interview Survey, United States, 2011–2013

Percentage Point Change
CharacteristicsAll, Avg Marginal EffectMale, Avg Marginal EffectFemale, Avg Marginal Effect
Labor market outcome and labor-related coverage measures
 Employed5.47.9*3.3
 In the labor force3.09.7**−2.9
 Employed full-time−0.5−2.01.9
 ESI−6.8*−6.5−6.2
 ESI offered7.9*5.311.4*
Insurance coverage and insurance-related measures
 Uninsured4.14.23.5
 Public0.5−0.72.2
 Private−4.2−3.7−4.7
 Directly purchased private insurance5.1*5.25.2
 Insurance coverage is worse15.4***13.317.7***

Note. ESI = employer-sponsored insurance. Estimates report the coefficient for T, a binary treatment variable equal to 1 if the respondent is at least 26 years old. In addition to the set of control variables, all regressions include age, age-squared, and their interactions with the treatment variable.

*P < .05; **P < .01; ***P < .001.

Summary statistics revealed differences in outcomes at the eligibility threshold (Table 2). Overall, and for men and women separately, being slightly older than the eligibility cutoff was associated with increases in having an offer of ESI coverage, being uninsured, having private health insurance, purchasing private insurance directly, and reporting that health insurance was worse than the previous year. Men aged 26 years or older were more likely to be employed and in the labor force than were their slightly younger counterparts who had the option of health coverage through a parent.

Table

TABLE 2— Definitions of Outcome Variables and Summary Statistics for Unmarried Young Adults: National Health Interview Survey, United States, 2011–2013

TABLE 2— Definitions of Outcome Variables and Summary Statistics for Unmarried Young Adults: National Health Interview Survey, United States, 2011–2013

All, Weighted Mean % (Unweighted No.)
Male, Weighted Mean % (Unweighted No.)
Female, Weighted Mean % (Unweighted No.)
Variable: DefinitionAge < 26 yAge ≥ 26 yAge < 26 yAge ≥ 26 yAge < 26 yAge ≥ 26 y
Employed: working for pay in the last 2 wk73.7 (5596)75.9* (4812)74.1 (2840)78.1** (2411)73.3 (2756)73.4 (2401)
In the labor force: working for pay or looking for work in the last 2 wk84.7 (5596)86.1 (4812)85.6 (2840)89.3*** (2411)83.7 (2756)82.5 (2401)
Employed full-time: working ≥ 32 h for pay in the last 2 wk74.0 (3992)78.0*** (3557)76.4 (2072)81.3** (1847)71.2 (1920)74.1 (1710)
Employer sponsored insurance: of the privately insured, covered through employer88.4 (3015)81.3*** (2205)87.4 (1608)82.0*** (1161)89.6 (1407)80.5*** (1044)
ESI offer: of those employed, share working for an employer offering health coverage59.8 (3983)64.7*** (3583)59.9 (2066)62.6 (1859)59.7 (1917)67.2*** (1724)
Uninsured: did not have health insurance coverage at the time of survey27.2 (5492)36.1*** (4730)30.7 (2789)41.7*** (2376)23.4 (2703)29.8*** (2354)
Public coverage: Medicaid, Medicare, or other public assistance/state sponsored plan13.3 (5492)13.8 (4730)7.6 (2789)7.2 (2376)19.7 (2703)21.3 (2354)
Private coverage: insurance provided in part or in full by an individual’s employer or union, or purchased directly by a person59.4 (5492)50.1*** (4730)61.7 (2789)51.2*** (2371)56.9 (2703)48.9*** (2354)
Direct purchase: private health coverage purchased directly, rather than through an employer or union5.4 (2896)10.7*** (2083)6.2 (1538)9.5** (1082)4.5 (1082)12.1*** (1001)
Worse insurance: compared with 1 y ago, health insurance coverage is worse (rather than the same or better)11.7 (2481)16.9*** (2358)10.8 (1195)15.1* (1103)12.6 (1286)18.7*** (1255)

Note. ESI = employer-sponsored insurance. Sample means are weighted. Sample sizes are in parentheses.

*P < .05;**P < .01;***P < .001.

I report the coefficient on the treatment () from estimating Equation 1 for each outcome. Regressions included the quadratic polynomial of age that fully interacted with a dichotomous indicator of treatment, individual-level characteristics, and year fixed effects, with SEs clustered at the individual level. Figures 1 through 3 show results before and after the young adult provision age threshold for each of the 10 outcomes, with quadratic fitted lines from the estimated parametric models (without controls) over the mean values of the share of individuals whose ages fall within the same 45-day age grouping.

The Overall Effect of the Young Adult Provision

Aging out of the provision had differential impacts by gender, but for the full sample there were changes in labor-related insurance outcomes, directly purchased private health insurance, and perceptions of health plan quality. At the threshold, the rates of employer-sponsored health insurance dropped (–6.8 percentage points), but shares of individuals with an offer of health insurance through an employer increased (+7.9 percentage points; Table 1). Graphically, these jumps are shown in Figure 1g and 1h.

The reduction in ESI is not entirely unexpected, as firms that offered health insurance during this timeframe had an average waiting period of approximately 2 months (after hiring) before ESI coverage began,15–17 and the NHIS question regarding ESI did not distinguish between coverage through the individual’s own employer or a parent’s employer.

Although there were no significant changes in type of health coverage (public, private, or uninsured), the 5.1-percentage-point uptick in directly purchased private health insurance suggested an interest in remaining insured. Figure 1i also shows this discrete change and demonstrates that after the increase at age 26 years, the probability of purchasing nongroup coverage initially rose and then began to decrease.

Importantly, I did not find that gains in insurance coverage resulting from the young adult provision eroded when individuals turned 26 years old, which again suggested that young adults wished to remain insured after losing eligibility through a parent. However, there was a 15.4-percentage-point increase in the probability of reporting that coverage was worse than 1 year previous. This finding was supported by the fact that directly purchased (i.e., non-ESI) health insurance typically provided less generous benefits than group (i.e., ESI) coverage.18 Furthermore, I showed an increase in directly purchased coverage once individuals aged out of the dependent coverage provision, as shown by the jump in dissatisfaction with health insurance plan is shown in Figure 1j. After the increase at age 26 years, the probability of reporting that coverage was worse than it was 1 year previously decreased.

When I estimated models for men and women separately, 2 different stories emerged: (1) labor market impacts are more direct for men, demonstrated by an increase in employment and labor force participation, and (2) they are more indirect for women, with an increase in employment-sponsored insurance being offered. Both suggested that the loss of eligibility for the young adult provision was associated with changes in labor and health coverage outcomes and increased reliance on the employer as a means of health insurance, or the potential return of job lock (i.e., remaining in a job in order to maintain health insurance coverage). Relaxing the employment–insurance connection suggested subsequent changes in employment choices and indicated that any reduction in job lock during the eligibility phase eroded on the 26th birthday.

The Impact on Young Men and Women

For young men, turning 26 years old led to a 7.9-percentage-point increase in employment and a nearly 10-percentage-point increase in labor force participation (Table 1). Figure 2a and 2b show the unadjusted jumps in labor market outcomes before and after the age of 26 years, which suggest that even 6 months, 1 year, 18 months, and 2 years past the eligibility threshold, employment and labor force participation rates were higher than they were during the eligibility period. These results implied that unmarried young men might have been using the ability to stay on a parent’s insurance as a reason to (temporarily) not participate in the labor market, or that job lock loosened during the eligibility phase.

Although men reentered the work force after losing eligibility for health insurance through a parent, there were no observed changes in labor force participation or employment for women. Among women at age 26 years, women reported an 11.4-percentage-point increase in offers by employers of insurance after aging out of the provision (Table 1). Figure 3h demonstrates a jump in ESI offer rates at 26 years and that after the jump, these rates decreased but did not return to the prethreshold levels, even 6 and 12 months after. There were no changes in the uninsured rate, so the increase in ESI offers suggested that women covered by parental insurance might have changed jobs at age 26 years to remain insured.

Turning 26 years old was associated with a nearly 18-percentage-point increase in women responding that their coverage was worse in that year versus the previous year (Table 1; Figure 3j). Because there were no significant changes in this measure for men, these results were possibly caused by the fact that women were higher utilizers of health care19 and might have had a health care experience under both types of insurance from which to judge coverage (e.g., annual pap and pelvic examinations).

I estimated models for different samples (wider age band, narrower age band, and inclusion of married individuals) and tested for model specification. Results from these tests suggested both appropriate model selection and accuracy of results (data available as a supplement to the online version of this article at http://www.ajph.org).

Using the NHIS, which contains birth date and interview date by month, I was able to identify loss of eligibility of parental insurance and to determine the immediate effects of aging out of the young adult provision of the ACA. Although existing literature demonstrated that the provision had many positive effects for the target population while eligible, to the best of my knowledge, this was the first analysis of how loss of eligibility altered individuals’ labor market and health coverage choices.

I found that ineligibility for the young adult provision did not lead to increases in uninsured rates, which suggested an interest in remaining insured, even in an era before the ACA’s individual health insurance mandate was in effect. Decisions made by young men and women on or around their 26th birthday demonstrated that provision ineligibility was associated with reliance on employment as a means of obtaining health insurance coverage or the potential return of job lock.

From a policy perspective, the differences in outcomes based on gender were particularly important. Results showed that men appeared to have been either more willing or more able (or both) to exit the labor force while eligible for the provision. This was supported by trends in living arrangements during the study period—in 2012, more than 1 in 3 young adults aged 18 to 31 years resided at home with their parents, with men being more likely than women to do so.20 Another explanation for the jump in employment and labor force participation rates among young men was that young women’s skills might have been better matched with their employment choices when the provision went into effect, which resulted in fewer of them leaving the labor force when becoming eligible. This might have been a carryover from the Great Recession (December 2007–June 2009), which was harder on men in terms of job loss during the economic contraction and job growth in the subsequent recovery.21

Placing the labor market findings from this article in context with literature examining the link between health insurance and labor supply decisions, it must be noted that the provision altered the health insurance choice set for eligible individuals. Before the provision, a young adult who desired to have health coverage had 3 primary options in the choice set: first, through an employer offering health insurance; second, through Medicaid or Medicaid-like program; and third, through directly purchased nongroup coverage. The provision gave individuals a fourth option—through a parent’s insurance plan. The provision is different from ESI and nongroup coverage because key requirements to receiving health insurance coverage, employment and (own) income are not present. However, it is somewhat similar to Medicaid coverage in that to receive health insurance, (own) employment is not required but dissimilar in its lack of an income restriction. Under the provision, labor force participation and income barriers were reduced, the value of parental insurance increased, and job lock loosened. The findings of decreased labor force participation and employment by young men were similar to those from a Medicaid study that demonstrated how an increase in the value of Medicaid led to reductions in labor force participation by single mothers.22

The exogenous change that reduced the possibility of job lock for eligible young adults might have had positive or negative societal well-being implications worth exploring in future research. If the young adults that temporarily exited the labor force used eligibility to improve their labor market match, then returned to the labor force in a career better suited to their job skill set, the provision could be viewed as welfare increasing. However, the provision might have been welfare decreasing if the individual simply exited the work force and returned to a job that required the same (or worse or fewer) skills as before exiting the market. In this case, the provision could be viewed as disrupting human capital formation, with longer-term effects potentially being observed over time (e.g., reduced lifetime earnings). Although there was not enough information in this data set to determine if individuals were using the provision to return to school, the period of time was associated with declining enrollment rates in graduate school programs,23 which suggested individuals were not substituting education for employment.

Although no statistically significant jumps in broad coverage type occurred at the threshold (public, private, or uninsured), changes within plan type (e.g., from a parent’s private insurance to their own private insurance) might have contributed to insurance coverage quality being perceived as worse than 1 year previously. For some, this might be a reflection of the first time that the young adult navigated the health care system on their own, and might not necessarily be a true indicator of plan quality.

Limitations

The NHIS is not without limitations. The public use data file did not contain state identifiers; therefore, I could not control for the fact that more than half of the states had already extended the age that young adults could remain on a parent’s health insurance plan when the ACA provision went into effect.24 However, Section 514 of the Employee Retirement Income Security Act of 1974 (ERISA)25 preempts state laws for self-insured plans,26 and during the study period, nearly 60% of private-sector employees with ESI had self-insured plans.27 In addition, most of these state-sponsored plans had stringent eligibility requirements to qualify for state coverage (e.g., unmarried, financially dependent on the parent, living in the same states as the parent, full-time student, younger than 25 years). For these reasons, it was plausible that my results found using national data were being driven by the federal law. Also, I used region of residence to control for geographic area, and other research found that most states did not experience a change in insurance coverage that was statistically different from the national change in insurance coverage as a result of the provision.6

Many colleges and universities mandate the purchase of health insurance,28 and thus, these individuals were more likely to be insured than nonstudents. For example, in the 2009–2010 school year, the overall rate of being uninsured for graduate and undergraduate students was 7.4%29 compared with rates of approximately 30% uninsured among the total population targeted by the dependent coverage provision.30 However, the NHIS did not include an indicator for student status, and although the ACA-dependent coverage provision extended coverage to all students up to age 26 years, many of these individuals were already insured before implementation. I used highest educational attainment to control for differences in education among young adults.

My sample was limited to unmarried individuals for several reasons outlined in the data section, but it is important to point out that marital status was a predictor of both labor and health insurance outcomes. To address this concern, I estimated models including married individuals, and I also controlled for marital status as part of the robustness checks (data available as a supplement to the online version of this article at http://www.ajph.org) to demonstrate that the inclusion of such individuals did not change the general findings.

Conclusions

My study focused on a period in time before the individual health insurance mandate began in effect, and so it would seem that moving forward there will be an increased interest in remaining insured at age 26 years. Many young adults will turn to state and federal health insurance marketplaces for information about health coverage. Because more than half of young adults (aged 18–29 years) regularly use 2 or more social media sites,31 marketplace education and outreach coordinators could use these sites to advertise to individuals getting ready to celebrate a 26th birthday. This is especially important for young men because, as my study demonstrated, they are more rapidly reentering the labor market and not necessarily selecting employment based on the potential offer of health insurance.

Human Participant Protection

This study was exempted from protocol approval by the institutional review board of The University of Minnesota because it does not constitute human participant research.

References

1. Kaiser Family Foundation. Employer health benefits, 2012 summary of findings. Available at: http://www.commonwealthfund.org/∼/media/Files/Publications/Issue%20Brief/2013/Aug/1701_Collins_covering_young_adults_tracking_brief_final_v4.pdf. Accessed November 26, 2014. Google Scholar
2. The Commonwealth Fund. Health insurance tracking survey of young adults. 2013. Available at: http://www.commonwealthfund.org. Accessed December 9, 2014. Google Scholar
3. Sommers BD, Buchmueller T, Decker SL, Carey C, Kronick R. The Affordable Care Act has led to significant gains in health insurance and access to care for young adults. Health Aff (Millwood). 2013;32(1):165174. Crossref, MedlineGoogle Scholar
4. Sommers BD, Kronick R. The Affordable Care Act and insurance coverage for young adults. JAMA. 2012;307(9):913914. Crossref, MedlineGoogle Scholar
5. Cantor JC, Monheit AC, Delia D, Lloyd K. Early impact of the Accordable Care Act on health insurance coverage of young adults. Health Serv Res. 2012;47(5):17731790. Crossref, MedlineGoogle Scholar
6. O’Hara B, Brault MW. The disparate impact of the ACA-dependent expansion across population subgroups. Health Serv Res. 2013;48(5):15811592. MedlineGoogle Scholar
7. Antwi YA, Moriya AS, Simon K. Effects of Federal Policy to Insure Young Adults: Evidence From the 2010 Affordable Care Act’s Dependent Coverage Mandate. Cambridge, MA: National Bureau of Economic Research; 2012. CrossrefGoogle Scholar
8. Minnesota Population Center and State Health Access Data Assistance Center. Integrated Health Interview Series: version 5.0. Minneapolis: University of Minnesota, 2012. Available at: http://www.ihis.us. Accessed June 13, 2015. Google Scholar
9. US Department of Labor. Young adults and the Affordable Care Act: protecting young adults and eliminating the burden on businesses and families. Available at: http://www.dol.gov/ebsa/faqs/faq-dependentcoverage.html. Accessed November 26, 2014. Google Scholar
10. US Census Bureau. Table 597. Labor force participation rates by marital status, sex, and age: 1970 to 2010. Available at: http://www.census.gov/compendia/statab/2012/tables/12s0598.pdf. Accessed December 24, 2014. Google Scholar
11. Bernstein AB, Cohen RA, Brett KM, Bush MA. Marital status is associated with health insurance coverage for working-age women at all income levels, 2007. NCHS Data Brief. 2008:(11)1–8. Google Scholar
12. US Census Bureau. Families and living arrangements, marital status, Table MS-2. Available at: http://www.census.gov/hhes/families/data/marital.html. Accessed November 26, 2014. Google Scholar
13. Carpenter C, Dobkin C. The effect of alcohol consumption on mortality: regression discontinuity evidence from the minimum drinking age. Am Econ J Appl Econ. 2009;1(1):164182. Crossref, MedlineGoogle Scholar
14. Yörük BK, Yörük CE. The impact of minimum legal drinking age laws on alcohol consumption, smoking, and marijuana use: evidence from a regression discontinuity design using exact date of birth. J Health Econ. 2011;30(4):740752. Crossref, MedlineGoogle Scholar
15. Kaiser Family Foundation. 2011 employer health benefits survey. Available at: http://kaiserfamilyfoundation.files.wordpress.com/2013/04/8225.pdf. Accessed December 3, 2014. Google Scholar
16. Kaiser Family Foundation. 2012 employer health benefits survey. Available at: http://kaiserfamilyfoundation.files.wordpress.com/2013/04/8345.pdf. Accessed December 3, 2014. Google Scholar
17. Kaiser Family Foundation. 2013 employer health benefits survey. Available at: https://kaiserfamilyfoundation.files.wordpress.com/2012/09/8465-employer-health-benefits-2013.pdf. Accessed December 3, 2014. Google Scholar
18. Reschovsky JD, Hadley J. The effect of tax credits for nongroup insurance on health spending by the uninsured. Health Aff (Millwood). 2004;W4-113W4-127. Google Scholar
19. Bertakis KD, Azari R, Helms JL, Callahan EJ, Robbins JA. Gender differences in the utilization of health care services. J Fam Pract. 2000;49(2):147152. MedlineGoogle Scholar
20. Fry R. A rising share of young adults live in their parents’ home. Pew Research Center. August 2013. Available at: http://www.pewsocialtrends.org/2013/08/01/a-rising-share-of-young-adults-live-in-their-parents-home. Accessed April 6, 2015. Google Scholar
21. Kochhar R. Two years of economic recovery: women lose jobs, men find them. Pew Research Center. July 2011. Available at: http://www.pewsocialtrends.org/2011/07/06/two-years-of-economic-recovery-women-lose-jobs-men-find-them. Accessed April 6, 2015. Google Scholar
22. Moffitt R, Wolfe B. The effect of the Medicaid program on welfare participation and labor supply. Rev Econ Stat. 1992;74(4):615626. CrossrefGoogle Scholar
23. The Council of Graduate Schools. Graduate schools see growth in applications and degrees, but enroll fewer new students in 2011. Available at: http://www.cgsnet.org/sites/default/files/E_and_D_2011_press_release_final.pdf. Accessed April 14, 2015. Google Scholar
24. National Conference of State Legislatures. Covering young adults through their parents’ or guardians’ health policy. Available at: http://www.ncsl.org/research/health/dependent-health-coverage-state-implementation.aspx. Accessed November 26, 2014. Google Scholar
25. Employee Retirement Income Security Act of 1974 (ERISA), Pub. L. No. 93-406, 88 Stat. 829 (codified as amended in scattered sections of 5 U.S.C., 18 U.S.C., 26 U.S.C., 29 U.S.C., and 42 U.S.C.). Google Scholar
26. Section 514 (29 U.S.C. § 1144). Google Scholar
27. Employee Benefit Research Institute. Self-insured health plans: state variation and recent trends by firm size. November 2012. Available at: http://www.ebri.org/pdf/notespdf/ebri_notes_11_nov-12.slf-insrd1.pdf. Accessed June 13, 2015. Google Scholar
28. American College Health Association National College Health Assessment. Undergraduate reference group executive summary, spring 2013. 2013. Available at: http://www.acha-ncha.org/docs/ACHA-NCHA-II_UNDERGRAD_ReferenceGroup_ExecutiveSummary_Spring2013.pdf. Accessed December 9, 2014. Google Scholar
29. American College Health Association. Undergraduate reference group executive summary, spring 2010. 2010. Available at: http://www.acha-ncha.org/docs/ACHA-NCHA-II_ReferenceGroup_ExecutiveSummary_Spring2010.pdf. Accessed December 24, 2014. Google Scholar
30. Centers for Medicare & Medicaid Services. Young adults and the Affordable Care Act. Protecting young adults and eliminating burdens on families and businesses. Available at: http://www.cms.gov/CCIIO/Resources/Files/adult_child_fact_sheet.html. Accessed December 9, 2014. Google Scholar
31. Duggan M, Ellison NB, Lampe C, Lenhart A, Madden M. Social media update 2014. Pew Research Center. January 2015. Available at: http://www.pewinternet.org/2015/01/09/social-media-update-2014. Accessed April 3, 2015. Google Scholar

Related

No related items

TOOLS

SHARE

ARTICLE CITATION

Heather M. Dahlen, MAAt the time of the study, Heather M. Dahlen was a PhD candidate with the Department of Applied Economics, University of Minnesota, Minneapolis. ““Aging Out” of Dependent Coverage and the Effects on US Labor Market and Health Insurance Choices”, American Journal of Public Health 105, no. S5 (November 1, 2015): pp. S640-S650.

https://doi.org/10.2105/AJPH.2015.302791

PMID: 26447916