Analysis of impacts of turning 26 (i.e., becoming ineligible for parental insurance coverage under the ACA young adult dependent coverage provision) on labor market and health insurance coverage.
Labor Market and Insurance Coverage Impacts Due to "Aging Out" of the Young Adult Provision of the Affordable Care Act
1. Labor Market and Insurance Coverage
Impacts Due to “Aging Out” of the Young
Adult Provision of the Affordable Care Act
Heather Dahlen
University of Minnesota, Applied Economics
2. The Young Adult Provision
Sept 23, 2010
Objective Data Methods Results Robustness Discussion
3. The Young Adult Provision
Sept 23, 2010
Allowed individuals to remain on a parent’s employer-
sponsored insurance (ESI) plan until age 26
Objective Data Methods Results Robustness Discussion
4. The Young Adult Provision
Sept 23, 2010
Allowed individuals to remain on a parent’s employer-
sponsored insurance (ESI) plan until age 26
Goal: Increase health insurance coverage for this group of
relatively healthy, previously uninsured individuals (which it
has)1
Objective Data Methods Results Robustness Discussion
1Sommers et. al (2012); Sommers and Kronick (2012); Cantor et. al (2012); O’Hara and Brault
(2013); Antwi, Mariya, and Simon (2012)
5. The Young Adult Provision
Gave young adults an alternative path to health
insurance coverage
Objective Data Methods Results Robustness Discussion
6. The Young Adult Provision
Gave young adults an alternative path to health
insurance coverage
By relaxing the tie between employment and health
insurance coverage, the employment/insurance
choice set was altered
Objective Data Methods Results Robustness Discussion
7. The Young Adult Provision
Gave young adults an alternative path to health
insurance coverage
By relaxing the tie between employment and health
insurance coverage, the employment/insurance
choice set was altered
− Potential reduction of job lock, or reliance on
employment for health insurance coverage
Objective Data Methods Results Robustness Discussion
8. How does aging out of the young adult
provision impact labor market and health
insurance coverage outcomes?
Objective Data Methods Results Robustness Discussion
9. National Health Interview Survey (NHIS)
Detailed information for a nationally representative sample
of non-institutionalized U.S. civilians
− Health
− Health insurance
− Employment
Objective Data Methods Results Robustness Discussion
10. National Health Interview Survey (NHIS)
Detailed information for a nationally representative sample
of non-institutionalized U.S. civilians
− Health
− Health insurance
− Employment
Accessed through the Integrated Health Interview Survey
(IHIS)
− Minnesota Population Center and State Health Access
Data Assistance Center
Objective Data Methods Results Robustness Discussion
11. Key Measures
Includes respondent birth month and year as well
as interview month and year
Able to more precisely account for time from 26th
birthday (eligibility threshold)
Objective Data Methods Results Robustness Discussion
12. Outcomes
Employment: Labor force participation, employed, and
full-time employment
Objective Data Methods Results Robustness Discussion
13. Outcomes
Employment: Labor force participation, employed, and
full-time employment
Employment-related health insurance measures:
Employer-sponsored insurance (ESI), offer of ESI
Objective Data Methods Results Robustness Discussion
14. Outcomes
Employment: Labor force participation, employed, and
full-time employment
Employment-related health insurance measures:
Employer-sponsored insurance (ESI), offer of ESI
Health Insurance: Plan quality compared to one year
prior, type of insurance (public, private, and
uninsured)
− Non-group directly purchased private coverage
Objective Data Methods Results Robustness Discussion
18. Sample
Years: 2011-2013
Ages: 24-28
N: 13,235
Subpopulations: Separate models based on gender
and marital status
Objective Data Methods Results Robustness Discussion
19. Objective Data Methods Results Robustness Discussion
Model
Regression Discontinuity (RD) design
Exploits the exogenous change in health coverage
options that occurs at the age cutoff for the young
adult provision program eligibility (age 26)
RD estimates the magnitude of the discontinuity in the
outcome at the cutoff
24. Objective Data Methods Results Robustness Discussion
RD estimates the
percentage point
change in an
outcome at age 26
25. Model
Logistic regressions
− Control for highest educational attainment, marital
status, region, health status, presence of a chronic
health condition, US citizenship, race/ethnicity, poverty,
gender (for full models), and year fixed effects
Objective Data Methods Results Robustness Discussion
26. Model
Treatment = 1 if age 26 or older
Age = distance from 26 (in months)
Objective Data Methods Results Robustness Discussion
27. Model
Treatment = 1 if age 26 or older
Age = distance from 26 (in months)
Objective Data Methods Results Robustness Discussion
29. Directly Purchased Private Insurance
4.4 pp increase (p<.05)
No other changes in
health insurance
coverage
Objective Data Methods Results Robustness Discussion
30. Directly Purchased Private Insurance
4.4 pp increase (p<.05)
No other changes in
health insurance
coverage
Prior to the individual
mandate
Objective Data Methods Results Robustness Discussion
31. Directly Purchased Private Insurance
4.4 pp increase (p<.05)
No other changes in
health insurance
coverage
Prior to the individual
mandate
Waiting periods for
employer-sponsored
insurance eligibility
Objective Data Methods Results Robustness Discussion
32. Insurance Coverage is Worse
(than 1 yr prior)
15.1 pp increase
Objective Data Methods Results Robustness Discussion
33. Insurance Coverage is Worse
(than 1 yr prior)
15.1 pp increase
First interaction
with the health
insurance on
own?
Objective Data Methods Results Robustness Discussion
34. Findings by Gender
Men
– At age 26: Increases in labor force participation (+7.5 pp) and
directly purchased nongroup insurance (+6.2 pp)
Objective Data Methods Results Robustness Discussion
35. Findings by Gender
Men
– At age 26: Increases in labor force participation (+7.5 pp) and
directly purchased nongroup insurance (+6.2 pp)
Interest in remaining insured
Were young men using the provision as a means of temporarily
exiting /delaying entry to the labor force?
Objective Data Methods Results Robustness Discussion
36. Findings by Gender
Men
– At age 26: Increases in labor force participation (+7.5 pp) and
directly purchased nongroup insurance (+6.2 pp)
Interest in remaining insured
Were young men using the provision as a means of temporarily
exiting /delaying entry to the labor force?
– Increases in health insurance coverage being worse (+12.2 pp)
Objective Data Methods Results Robustness Discussion
37. Findings by Gender
Men
– At age 26: Increases in labor force participation (+7.5 pp) and
directly purchased nongroup insurance (+6.2 pp)
Interest in remaining insured
Were young men using the provision as a means of temporarily
exiting /delaying entry to the labor force?
– Increases in health insurance coverage being worse (+12.2 pp)
Women
– Large increase (+17.6 pp) in reporting of insurance coverage being
worse one year prior
Higher healthcare utilization rates
Objective Data Methods Results Robustness Discussion
38. Findings for Unmarried Individuals
Men
– Increase in employment (+7.9 pp)
– Increase in labor force participation (+9.7 pp)
Objective Data Methods Results Robustness Discussion
39. Findings for Unmarried Individuals
Men
– Increase in employment (+7.9 pp)
– Increase in labor force participation (+9.7 pp)
Women
– Increase in employer-sponsored insurance offers (+11.7pp)
– Increase in health coverage being worse (+17.7 pp)
Objective Data Methods Results Robustness Discussion
40. Model Specification and Robustness Checks
1. Smoothness of the model covariates
No significant jumps at age 26
Objective Data Methods Results Robustness Discussion
41. Model Specification and Robustness Checks
1. Smoothness of the model covariates
No significant jumps at age 26
2. Respondent should not have control over the forcing
variable (the cut-point)
Age is the forcing variable and this is naturally satisfied
Objective Data Methods Results Robustness Discussion
42. Model Specification and Robustness Checks
1. Smoothness of the model covariates
No significant jumps at age 26
2. Respondent should not have control over the forcing
variable (the cut-point)
Age is the forcing variable and this is naturally satisfied
3. No non-random sorting to one side of the threshold
Plotted the distribution of young adults around the eligibility
threshold and this did not occur
Objective Data Methods Results Robustness Discussion
43. Model Specification and Robustness Checks
4. Model Fit. Estimated models for the following:
- A) Same age primary sample but earlier years (2004-2006):
No significant results
Objective Data Methods Results Robustness Discussion
44. Model Specification and Robustness Checks
4. Model Fit. Estimated models for the following:
- A) Same age primary sample but earlier years (2004-2006):
No significant results
- B) Only individuals younger than 26, same years as primary
study (2011-2013), and artificial eligibility threshold: No
significant results
Objective Data Methods Results Robustness Discussion
45. Model Specification and Robustness Checks
4. Model Fit. Estimated models for the following:
- A) Same age primary sample but earlier years (2004-2006):
No significant results
- B) Only individuals younger than 26, same years as primary
study (2011-2013), and artificial eligibility threshold: No
significant results
- C) Only individuals older than 26, same years as primary
study (2011-2013), and artificial eligibility threshold: No
significant results
Objective Data Methods Results Robustness Discussion
46. Model Specification and Robustness Checks
4. Model Fit. Estimated models for the following:
- A) Same age primary sample but earlier years (2004-2006):
No significant results
- B) Only individuals younger than 26, same years as primary
study (2011-2013), and artificial eligibility threshold: No
significant results
- C) Only individuals older than 26, same years as primary
study (2011-2013), and artificial eligibility threshold: No
significant results
Objective Data Methods Results Robustness Discussion
47. Model Specification and Robustness Checks
5. Sample Appropriateness. Estimated the following models:
- A) Narrower age band: results are less precise
- B) Wider age band: includes individuals further removed from
the threshold and have had more time to adjust (however,
many of the significant results from primary models remain)
- C) Restriction to unmarried
Objective Data Methods Results Robustness Discussion
48. First analysis of how loss of eligibility for the young adult
provision alters labor market and health coverage choices
Objective Data Methods Results Robustness Discussion
49. First analysis of how loss of eligibility for the young adult
provision alters labor market and health coverage choices
No change in uninsurance rate + increase in directly
purchased coverage = young adults are interested in
remaining insured
Objective Data Methods Results Robustness Discussion
50. First analysis of how loss of eligibility for the young adult
provision alters labor market and health coverage choices
No change in uninsurance rate + increase in directly
purchased coverage = young adults are interested in
remaining insured
Larger labor market effects for unmarried men and women
Objective Data Methods Results Robustness Discussion
51. First analysis of how loss of eligibility for the young adult
provision alters labor market and health coverage choices
No change in uninsurance rate + increase in directly
purchased coverage = young adults are interested in
remaining insured
Larger labor market effects for unmarried men and women
Increase in labor force participation for young men
– Graduate school enrollment rates did not increase
during this time
Objective Data Methods Results Robustness Discussion
52. Large jumps in health insurance plan dissatisfaction at age
26
Objective Data Methods Results Robustness Discussion
53. Large jumps in health insurance plan dissatisfaction at age
26
– Health insurance marketplace education and outreach
coordinators can use the results for targeted marketing of
young adults nearing a 26th birthday
− Smooth the coverage transition and reduce plan quality
dissatisfaction
Objective Data Methods Results Robustness Discussion