The Affordable Care Act (ACA) extended health insurance coverage to millions of Americans through Medicaid expansions in participating states, new regulations to make private insurance coverage accessible to all applicants, and federal tax credits for low-to-moderate-income individuals without affordable employer coverage. The law also created new inducements to enroll in insurance by imposing penalties on most individuals who remain uninsured. Previous research on the effects of the ACA on health insurance has found large gains in insurance coverage since the main provisions of the law took effect in January 2014. For example, using data from the Gallup-Healthways Well-Being Index, Sommers et al. (2014) reported that a net 10.3 million adults gained health insurance between the last quarter of 2013 and the second quarter of 2014. A similar study fielded by the Commonwealth Fund (Collins, Rasmussen, and Doty, 2014) found that a net 9.5 million adults gained insurance between the third quarter of 2013 and the second quarter of 2014. However, to date, there is limited information about what types of coverage the newly insured have obtained and the insurance history of those who enrolled in coverage through Medicaid or the Marketplaces. In this report, we track a fixed population of individuals over time to understand how their insurance status has changed since the major provisions of the ACA went into effect at the beginning of 2014. Our study builds on previous research by using a longitudinal design to estimate not only overall gains in insurance coverage, but also transitions in insurance coverage. Starting with a cohort of people who were surveyed in September 2013, we track changes in insurance enrollment through November 2014. Our design enables us to observe detailed information about insurance transitions, such as what types of insurance the newly insured have gained and what types of coverage people have lost. This allows us to observe not only enrollment in new sources of insurance, such as the Marketplaces and Medicaid (for single adults in expansion states), but also gains in other sources of insurance, including employer-sponsored insurance (ESI).
The experiences of those who are insured for the first time may be very different from those who were previously insured through other sources. The U.S. Department of Health and Human Services (HHS, 2014) reported that 87 percent of individuals enrolling in coverage through a federally facilitated Marketplace were uninsured at the time of application. However, this figure may overestimate the number of Marketplace enrollees who were previously uninsured because some may have been uninsured for only a very brief period of time before enrolling (such as those who purposefully dropped other coverage to become enrolled on the Marketplaces or whose plan was terminated because it did not meet ACA requirements).
We tracked transitions in insurance coverage using the RAND Health Reform Opinion Study (RHROS), a longitudinal survey that followed a cohort of people from September 2013 through November 2014. While our analysis focused on transitions between September 2013 and November 2014, the month immediately before the ACA's first open enrollment period and again immediately before the ACA's second open enrollment period, we also looked at transitions over the entire period.
The RHROS is conducted using the RAND American Life Panel (ALP), a nationally representative panel of individuals who regularly participate in surveys. The RHROS has been conducting monthly surveys with the panel between November 2013 and May 2014 and then quarterly since then, contacting the same group of individuals each month. In addition to asking respondents about their opinions of the ACA, each month we collected information about enrollment in health insurance, including ESI, Medicaid, Medicare, insurance purchased on a Marketplace, and other insurance purchased on the individual market. We can identify the health insurance status of RHROS respondents in September 2013 by linking them to data previously collected through the ALP, allowing us to estimate the number of individuals transitioning from one source of coverage to another. RAND's Human Subjects Protection Board has approved the ALP, all data collection, and the study design reported here. Participation in the ALP, including response to each question, is optional.
Our analyses were limited to respondents recruited through probability-based sampling, such as random-digit dialing and address-based sampling.
Our analysis is based on a sample of 1,628 respondents between the ages of 18 and 64 who provided information about their insurance coverage between August 23, 2013, and September 30, 2013 (for simplicity we refer this as September 2013 in the remainder of the report), and again between November 11, 2014, and December 1, 2014 (referred to as November 2014 in the remainder of the report). Our sample is limited to ALP respondents recruited using a probability-based sampling method (2,953) who responded in both 2013 and 2014 (2,329 in 2013, 1,972 in 2014, and 1,636 in both) and provided usable* information about their source of insurance (leaving 1,628). Individuals are drawn from a household population, excluding people in institutionalized settings, such as nursing homes.
Using our sample, we estimated the number of people in the U.S. population as a whole in each insurance category, as discussed in more detail below. We used sample weights so that our sample is representative of the population, benchmarking key demographic characteristics to the 2013 Current Population Survey (CPS), a large national survey conducted by the U.S. Census Bureau and the U.S. Bureau of Labor Statistics. We aimed to match population proportions on interactions of gender and race/ethnicity, gender and education, gender and age, and household income interacted with household size. However, we did not include health insurance status. Furthermore, we adjusted our weights to address nonresponse to the 2014 survey among those responding in 2013. We then used the weighted percentage of respondents from our survey multiplied by the total U.S. population between the ages of 18 and 64 (198.6 million) to extrapolate to the national level (U.S. Census Bureau, 2014). We reported the margin of error based on a 95-percent confidence interval.** A comparison of weighted characteristics of our sample and that of the CPS and more detailed information about the methods are available in our companion methodology paper, Carman and Eibner (2015).
Table 1 highlights the net gains and losses in coverage. We estimate that, on net, 10.8 million people gained coverage through Medicaid, 7.6 million people gained coverage in the Marketplaces, and 6.6 million people gained coverage through ESI between September 2013 and November 2014. Coverage through self-pay policies and through other sources (such as Medicare, military insurance, and other state policies) declined by 1.3 million and 10.7 million, respectively. A number of factors contribute to the large drop in other sources of coverage, including the elimination of small state safety net programs that coincided with the increase in Medicaid eligibility. In total, 12.9 million additional people were insured during the study period; the number of uninsured people declined from 40.2 million in September 2013 to 27.3 million in November 2014.
Table 1. Net Changes in Insurance Coverage, September 2013 to November 2014
|ESI||115.3 (+/– 7.6)||121.9 (+/– 7.4)||6.6 (+/– 5.9)|
|Medicaid||10.4 (+/– 2.5)||21.2 (+/– 4.1)||10.8 (+/– 3.8)|
|Self-pay||8.5 (+/– 2.6)||7.2 (+/– 2.2)||(1.3) (+/– 2.3)|
|Marketplace||— —||7.6 (+/– 2.4)||7.6 (+/– 2.4)|
|Other||24.0 (+/– 5.9)||13.3 (+/– 4.6)||(10.7) (+/– 4.6)|
|Subtotal: insured||158.3 (+/– 6.3)||171.3 (+/– 5.4)||12.9 (+/– 6.0)|
|Uninsured||40.2 (+/– 6.3)||27.3 (+/– 5.4)||(12.9) (+/– 6.0)|
NOTES: Numbers in italics show margins of error. Margin of error represents a 95-percent confidence interval.
In Table 2, we show the number of people transitioning from uninsured to insured status, or vice versa. The net 12.9-million-person increase in insurance was driven by 20.4 million adults ages 18 to 64 who gained coverage during the study period and 7.4 million people who lost coverage. There were 19.8 million people who were uninsured in both September 2013 and November 2014 and 150.9 million who were insured in both time periods.
Table 2. Transitions in Insurance Coverage from September 2013 to November 2014
|Uninsured||19.8 (+/– 4.71)||20.4 (+/– 6.38)||40.2 (+/– 4.90)|
|Insured||7.4 (+/– 3.31)||150.9 (+/– 6.81)||158.3 (+/– 6.38)|
|Total||27.3 (+/– 5.53)||171.3 (+/– 5.53)||198.5|
NOTES: All numbers (including margin of error) are in millions of individuals. Light gray cells show categories that did not change from 2013 to 2014 (i.e., individuals who experienced no transition). Dark gray cells show numbers of transitions from 2013 to 2014. Numbers in italics show margins of error. Margin of error represents a 95-percent confidence interval.
Table 3 provides detailed transitions between type of insurance coverage in 2013 and 2014. Among those gaining coverage, most (7.5 million) enrolled in Medicaid, followed by employer plans (7.3 million), the Marketplaces (3.1 million), non-Marketplace private plans (1.5 million), and all other insurance sources (1.0 million). Among those starting out with insurance, 3.4 million people transitioned from employer coverage to uninsured status, 0.5 million transitioned from Medicaid to uninsured status, and 2.8 million transitioned from other sources of coverage to uninsured status. Most respondents (86 percent) experienced no change in insurance status during the study period, either remaining insured or uninsured over the period. Of the 158.3 million people with insurance in September 2013, the vast majority (81 percent) experienced no changes in the source of their insurance during the study period.*** Among those who were uninsured at baseline, just under half (49 percent) remained uninsured at follow-up.
Table 3. Transitions Across Insurance Categories from September 2013 to November 2014
|No Insurance||ESI||Medicaid||Individual Market
|No Insurance||19.8||7.3 (+/– 3.77)||7.5 (+/– 2.88)||1.5 (+/– 1.05)||3.1 (+/– 1.32)||1.0 (+/– 0.74)||40.2 (+/– 6.38)|
|ESI||3.4 (+/– 1.71)||105.9 (+/– 7.70)||1.5 (+/– 1.25)||1.1 (+/– 0.66)||2.2 (+/– 1.63)||1.2 (+/– 0.97)||115.3 (+/– 7.70)|
|Medicaid||0.5 (+/– 0.36)||0.8 (+/– 0.70)||7.7 (+/– 2.18)||—||0.5 (+/– 0.66)||1.0 (+/– 0.70)||10.4 (+/– 2.53)|
|Individual Market||0.7 (+/– 0.70)||2.3 (+/– 1.63)||0.0 (+/– 0.06)||4.5 (+/– 1.79)||0.9 (+/– 0.70)||0.1 (+/– 0.13)||8.5 (+/– 2.61)|
|Other||2.8 (+/– 2.72)||5.6 (+/– 3.00)||4.0 (+/– 1.83)||0.2 (+/– 0.20)||0.9 (+/– 0.74)||10.1 (+/– 4.51)||24.0 (+/– 5.99)|
|Total||27.3 (+/– 5.53)||121.9 (+/– 7.47)||21.2 (+/– 4.16)||7.2 (+/– 2.18)||7.6 (+/– 2.41)||13.3 (+/– 4.67)||198.5|
NOTES: All numbers (including margin of error) are in millions of individuals. Light gray cells show numbers that did not change from 2013 to 2014 (i.e., individuals who experienced no transition). Numbers in italics reflect margins of error. Margin of error represents a 95-percent confidence interval.
Table 3 also shows the prior insurance status of Marketplace enrollees. We estimated that 7.6 million people had Marketplace coverage in November 2014, a number that is close to the federal government's revised administrative tally of 6.9 million enrollees. Of those 7.6 million individuals estimated to have enrolled in Marketplace coverage based on our data, 3.1 million (41 percent) were uninsured in September 2013.
We also investigated monthly trends in insurance enrollment. Figure 1 illustrates the number of people with each insurance type in September 2013, followed by the number in each category for various time points reflecting coverage in 2014. In November 2013 and December 2013, respondents were asked about their insurance coverage for 2014. In the 2014 surveys, respondents were asked about current coverage. Figure 1 illustrates that the percentage of respondents with insurance coverage grew consistently from November 2013 through May 2014, with gains occurring in ESI, Medicaid, Marketplaces, and other and decreases in non-Marketplace individual market coverage. Among those who purchased insurance on the Marketplaces, we observe the most growth in April and May 2014, consistent with the surge in enrollment reported by HHS. The number enrolled in Marketplace coverage approximately doubled between November 2013 and November 2014. Between May and August, there was a slight rise in the uninsured rate, although these changes are within the margin of error. No single source of coverage had any significant change in coverage between May 2014 and November 2014.
One result that merited further analysis was the number of people gaining coverage through ESI. We investigated changes in access to ESI that may have been related to this trend. In May 2014 respondents were also asked about their current access to insurance coverage, as well as their access in September 2013. Figure 2 illustrates the share with access to ESI for three key groups. First, among those who were uninsured in both years, only 18 percent reported having access to ESI in 2013. Second, and most notably, among those who were uninsured in 2013 and had ESI in 2014, 43 percent had access to ESI in 2013. In other words, of those newly covered by ESI, 57 percent did not previously have access to ESI. Finally, in the overall population, 68 percent reported access to ESI in 2013, 7 percent of whom did not enroll in ESI. Results for access and take-up of ESI offers in 2014 were not statistically significantly different from those from 2013.
These data provide a unique opportunity to study insurance transitions across a broad representative sample of the United States, but there are some limitations. First, the sample size is limited to only 1,628 observations. The small sample size reduces the precision of our estimates and leads to relatively wide margins of error. For example, our estimate that approximately 12.9 million people gained insurance coverage between 2013 and 2014 has a margin of error of 6 million. Second, respondents are asked to report their own insurance status. Some respondents may not correctly report the type of insurance coverage they have; in particular, people may have difficulty distinguishing Marketplace coverage from other nongroup coverage (Pascale et al., 2013). It is also possible that people may mix up Marketplace and Medicaid coverage, especially because the ACA's “no wrong door” policy means that some people who attempt to enroll through a Marketplace website may ultimately be transferred to Medicaid. Third, one concern with panel data is that participation in later waves may be influenced by the variables of interest—in this case, that insurance choices may influence the decision to participate in later waves of the survey. To address this concern, our survey weights adjust for nonresponse associated with factors that are observable in our data. A strength of the longitudinal approach is that it avoids recall bias that may occur when respondents are asked to retrospectively report about prior insurance coverage.
We found that 20.4 million people became newly insured and 7.4 million become uninsured between September 2013 and November 2014, for a net gain in insurance of 12.9 million among adults ages 18 to 64. These estimates correspond to a reduction in the uninsurance rate among this age group from 20.3 percent to 13.7 percent. The net gains that we estimated are similar to those estimated in prior cross-sectional studies (Sommers et al., 2014; Collins et al., 2014; and Long et al., 2014), with slight differences due to sampling variability, differences in the way questions were worded, and differences in the dates of the surveys.
Among those newly insured, most enrolled in Medicaid, followed by ESI and the Marketplaces. Just over 40 percent of Marketplace enrollees were uninsured in September 2013; the rest transitioned from other coverage sources, including employer, nongroup, and all others. Our estimate of the share of Marketplace enrollees who were previously uninsured is lower than HHS's estimate, which focused on uninsurance at the time of Marketplace application. Some individuals who were uninsured at the time of application may have lost or dropped coverage very recently, perhaps with the intention of switching to the Marketplaces.
Our data suggest that a large share of people gaining insurance became newly enrolled in employer coverage. While this may seem surprising, given the intense media attention focused on the Marketplaces, employer coverage is by far the largest source of insurance among Americans under age 65, and the ACA creates new incentives for people to take up employer policies. Specifically, while the ACA mandates that most individuals must enroll in insurance, people are generally ineligible for Marketplace subsidies if they have an affordable offer of coverage from their employer. Gains in ESI were also found following Massachusetts' health reform (Long and Stockly, 2010, and Kenney, Long, and Luque, 2010). Our analysis suggests that 43 percent of the individuals who transitioned from uninsured status to employer coverage did so by taking up an offer that was previously available. The remainder became newly eligible for employer coverage. However, we cannot determine whether these new offers were due to changes in job status, changes in employer offering behavior, or other factors. Furthermore, our results find no significant change in access to ESI from 2013 to 2014, in contrast to the concern that employers would drop ESI.
While the vast majority (81 percent) of those previously insured experienced no change in their source of coverage, 7.4 million people lost coverage over the period studied, and 22.7 million moved from one source of coverage to another. Transitions in health insurance coverage are common in the United States, and they occur for a variety of reasons, including losing or gaining employment, family transitions, and aging in and out of eligibility for certain programs (Graves and Swartz, 2013). Recent estimates suggest that the share of people losing coverage between 2013 and 2014 was no higher than the share of people who lost coverage in prior years. For example, using the Medical Expenditure Panel Survey, Vistnes, and Cohen (2015) report that 6 percent of nonelderly adults who were insured in 2012 lost coverage by 2013, compared with 5.5 percent of nonelderly adults losing coverage between 2013 and 2014. Our data suggest that 3.8 percent of nonelderly adults who were insured in 2013 lost coverage in 2014.
One concern frequently cited by public officials and the media was that people may have lost individual market coverage due to plan cancelations. We found that, although enrollment in non-Marketplace individual market coverage declined, the total enrolled in any type of individual coverage (including on or off the Marketplaces) increased from 8.5 to 14.8 million. The vast majority of those with individual market insurance in 2013 remained insured in 2014, suggesting that—even among those who had their individual market policies canceled—most found coverage through an alternative source. Others who had their policies canceled may have become eligible for the ACA's tax credits, potentially making the ACA's Marketplace plans more affordable than their previous nongroup policies.
The ACA has greatly expanded health insurance coverage in the United States, with minimal effect on those who were insured before the major provisions of the law took effect. In addition, the law has expanded coverage using all parts of the health insurance system, including employer plans, Medicaid, and the newly created Marketplaces. Understanding these transitions is critical to assess the effects of the law and to determining how many people may need assistance as they gain access to new or new types of insurance. In particular, those who became insured for the first time may need help navigating the system, knowing what questions to ask, and determining when to seek care. Those who became insured through the Marketplaces may face higher deductibles than other patients or face smaller provider networks and thus may need help making cost-conscious choices about the type and location of care without sacrificing quality.
Carman KG and Eibner C, Methodology of the RAND Health Reform Opinion Study, Santa Monica, Calif.: RAND Corporation, RR-947-RC, 2015. As of April 29, 2015:
Collins SR, Rasmussen PW, and Doty MM, “Gaining Ground: American's Health Insurance Coverage and Access to Care After the Affordable Care Act's First Open Enrollment Period,” The Commonwealth Fund, 2014. As of July 24, 2014:
Graves JA and Swartz K, “Understanding State Variation in Health Insurance Dynamics Can Help Tailor Enrollment Strategies for ACA Expansion,” Health Affairs, Vol. 32, No. 10, 2013, pp. 1832–1840.
HHS— See U.S. Department of Health and Human Services.
Kenney G, Long SK, and Luque A, “Health Reform in Massachusetts Cuts the Uninsurance Rate for Children in Half,” Health Affairs, Vol. 29, No. 6, 2010, pp. 1242–1247.
Long SK, Denney GM, Zuckerman S, et al., “QuickTake: Number of Uninsured Adults Continues to Fall Under the ACA: Down by 8.0 million in June 2014,” Urban Institute, Health Policy Center, Health Reform Monitoring Survey, 2014. As of July 24, 2014:
Long SK and Stockley K, “Sustaining Health Reform in a Recession: An Update on Massachusetts as of Fall 2009,” Health Affairs, Vol. 29, No. 6, 2010, pp. 1234–1241.
Pascale J, Rodean J, Leeman J, Cosenza C, and Schoua-Glusberg A, “Preparing to Measure Health Coverage in Federal Surveys Post-Reform: Lessons from Massachusetts,” Inquiry, Vol. 50, No. 2, May 2013, pp. 106–123.
Sommers BD, Musco T, Finegold K, et al., “Health Reform and Changes in Health Insurance Coverage in 2014,” New England Journal of Medicine, Vol. 371, 2014, pp. 867–874. As of July 24, 2014:
U.S. Census Bureau, “State and County Quickfacts,” U.S. Department of Commerce, March 27, 2014. As of April 7, 2014:
U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, “Health Insurance Marketplace: Summary Enrollment Report for the Initial Annual Open Enrollment Period,” ASPE Issue Brief, May 1, 2014. As of July 24, 2014:
Vistnes J and Cohen SB, “Transitions in Health Insurance Coverage over Time, 2012–2014 (Selected Intervals): Estimates for the U.S. Civilian Noninstitutionalized Adult Population Under Age 65,” Agency for Healthcare Research and Quality, Medical Expenditure Panel Survey, Statistical Brief #467, February 2015.
* In a few cases, respondents provided a write-in response to our insurance question that was unclear and could not be used. For example, if someone listed “Aetna” as his or her insurer without additional information, we could not be sure whether the individual was enrolled in exchange coverage, other nongroup coverage, or an employer plan.
** We estimated the margin of error using plus or minus one-half the width of the 95-percent confidence interval.
*** Per Table 3, 105.9 million people retained employer coverage, 7.7 million people retained Medicaid coverage, 4.5 million retained individual market coverage, and 10.1 million retained other coverage, for a total of 128.2 million people retaining the same coverage between 2014 and 2015. By comparing that total with the 158.3 million people with insurance in 2013, we can infer that 81 percent (128.2/158.3) had no change in insurance status.
The research described in this article was conducted by RAND Health.