In 2014 the Affordable Care Act (ACA)

Transcription

1 By John H. Goddeeris, Stacey McMorrow, and Genevieve M. Kenney DATAWATCH Off-Marketplace Enrollment Remains An Important Part Of Health Insurance Under The ACA The introduction of Marketplaces under the Affordable Care Act greatly expanded individual-market health insurance coverage in 2014, but millions of adults continued to purchase individual coverage outside of the Marketplaces. They were more likely to be male, be white, have higher incomes, and be in excellent or very good health, compared to Marketplace enrollees. doi: /hlthaff HEALTH AFFAIRS 36, NO. 8 (2017): Project HOPE The People-to-People Health Foundation, Inc. In 2014 the Affordable Care Act (ACA) made health insurance more accessible to nonelderly adults without employersponsored coverage or Medicaid by creating highly regulated federal and state-based Marketplaces, through which eligible individuals could purchase coverage with federal subsidies. 1 Although details about the people who enrolled in individual Marketplace coverage have steadily emerged, 2 4 relatively little is known about either those who purchased coverage in the individual (nongroup) market outside of the Marketplaces after the ACA s implementation or how the individual market as a whole has changed. Using data from the National Health Interview Survey (NHIS), we found that the number of adults ages covered by individual-market insurance nearly doubled after the ACA s implementation, increasing from 7.7 million in 2013 to 14.9 million in 2015 (Exhibit 1). There were 8.6 million nonelderly adult enrollees in Marketplace plans in 2015, and 6.3 million enrollees in off-marketplace plans. The share of lowerincome enrollees in the individual market as a whole grew, driven by Marketplace enrollment, while off-marketplace enrollees had higher incomes (Exhibit 2). While lawmakers continue to debate changes to the ACA, any modification or alternative is likely to rely heavily on the individual market in one form or another. Our findings provide John H. Goddeeris (goddeeri@ msu.edu) is a professor of economics at Michigan State University, in East Lansing, and an affiliated scholar of thehealthpolicycenter, Urban Institute, in Washington D.C. Stacey McMorrow is a senior research associate at the Health Policy Center, Urban Institute. Genevieve M. Kenney is codirector of and a senior fellow at the Health Policy Center, Urban Institute. Exhibit 1 Adults ages enrolled in the individual health insurance market in 2013 and 2015, by income and market segment SOURCE Authors analysis of data for 2013 and 2015 from the National Health Interview Survey. NOTES The individual market refers to all directly purchased private health insurance, including both Marketplace and off-marketplace plans in FPL is federal poverty level. August :8 Health Affairs 1489

2 Exhibit 2 Income distribution of individual-market enrollees ages 19 64, 2013 and 2015 SOURCE Authors analysis of data for 2013 and 2015 from the National Health Interview Survey. NOTES The individual market is explained in Exhibit 1 Notes. All differences between 2013 and 2015 individual-market enrollees, and all differences between enrollees in Marketplace and those in off-marketplace plans, are significant (p < 0:05), except for income of percent of the federal poverty level (FPL). No differences between 2013 individual-market enrollees and 2015 enrollees in off-marketplace plans are significant (p < 0:05). important context for contemplating the future of that market. Study Data And Methods Study Population We focused on the nonelderly adult population (ages 19 64), the primary targets of the ACA coverage expansions, and we used NHIS data for the period We examined the characteristics of adults purchasing coverage in the individual market in the year before and the two years after ACA implementation, and we also compared purchasers of Marketplace and off-marketplace plans. We classified families in the NHIS data into health insurance units, which more closely resemble the units used to estimate income eligibility for tax credits or Medicaid, and we calculated the income of health insurance units relative to the federal poverty level. 6 We defined individual-market enrollees as survey respondents who reported having directly purchased private insurance. To identify the subset of enrollees in Marketplace plans, we used a variable provided by the National Center for Health Statistics. 7 We classified the remaining individual-market enrollees as purchasers of off-marketplace plans. All NHIS interviews are conducted in person, and information is solicited about specific plan names, with follow-up questions to minimize errors in classification of coverage. 8 We examined demographic and socioeconomic characteristics of enrollees in the individual market, as well as measures of self-reported general health status, current smoking and obesity, and reports of ever having been diagnosed with hypertension or diabetes. We defined obese adults as people with a body mass index (BMI) greater than 30 kg/m 2, and smokers as people who reported smoking every day or some days. All health status measures except general health status were available for only one sampled adult per family. We also measured continuity and quality of coverage by identifying individualmarket enrollees who had been uninsured at any time in the twelve months before their NHIS interview, and those who reported that their coverage was better than it had been one year prior to their interview. Analyses All analyses used NHIS survey weights, and standard errors were adjusted to account for the complex survey design. We used chi-square tests to assess differences in the distributions of respondents across age, racial/ ethnic, and general health status groups. We used two-sided t-tests to assess differences in the means of other characteristics across groups. 9 Limitations This study had several limitations. First, all information was self-reported and therefore subject to measurement error of various forms, including recall biases. Second, despite efforts by the National Center for Health Statistics, which administers the survey, to assess coverage types accurately, 8 insurance coverage is likely to be misclassified for some people, and we were unable to distinguish between ACA-compliant and noncompliant plans outside the Marketplace. Third, we restricted our analysis to nonelderly adults ages 19 64, so the NHIS enrollment numbers presented here for the individual market and Marketplace are noticeably smaller than those reported using administrative data sources on enrollees of all ages. For example, data submitted by insurers to the Department of Health and Human Services to comply with medical loss ratio (MLR) regulations under the ACA use total enrollments that combine adults and children. However, when we included children up to age eighteen in our NHIS estimates, we found that NHIS and MLR data provided very similar views of the size of the individual market (see online Appendix Table A1). 10 When children are included, NHIS estimates are also generally consistent with administrative data on Marketplace enrollment (Appendix Tables A2 and A3) Health Affairs August :8

3 Fourth, there were three main sources of error in our income measure. First, income and earnings details were imputed by the NCHS for approximately 25 percent of the sample. Second, we redistributed family income to health insurance units that make up a family, but not all sources of income could be attributed to individual family members. Third, the NHIS collects information on income for the previous calendar year, while eligibility for various programs depends on current income. Finally, our analysis described the composition of the individual market before and after the ACA s implementation, but it did not allow us to isolate the effects of the law from those of other changes that were occurring at the same time. Study Results The number of adults ages covered by individual-market insurance increased from 7.7 million in 2013 to 14.9 million in 2015, as noted above (Exhibit 1), and the share of lower-income enrollees in the market as a whole increased (Exhibit 2). Individual-market enrollment expanded for all subgroups that we examined between 2013 and 2015, but the relative gains in enrollment differed across groups, and the composition of the market varied a great Exhibit 3 Selected characteristics of individual-market enrollees ages 19 64, 2013 and 2015 Number of enrollees (1,000s) 2013 a 2015 b enrollees, 2013 a All individual-market Distribution of enrollees by age, sex, income, race/ethnicity, education, citizenship, and work and marital status Off-Marketplace enrollees, 2015 c Age group (years) , % 14.0% 10.9% ,302 2, ,310 2, ,887 3, ,244 4, Sex Female 4,002 7, Income (percent of poverty) 100% or less 834 1, % 138% 358 1, % 250% 1,301 3, % 400% 1,681 3, More than 400% 3,562 4, Race/ethnicity White, non-hispanic 6,196 9, Black, non-hispanic 301 1, Other, non-hispanic 701 1, Hispanic 538 2, Education Less than high school 294 1, High school graduate 1,519 3, Some college 2,457 4, College graduate 3,466 5, Citizenship status Noncitizen 449 1, Work status Full-time worker 4,285 8, Marital status Married 4,469 7, Marketplace enrollees, 2015 d SOURCE Authors analysis of data for 2013 and 2015 from the National Health Interview Survey. NOTES The individual market is explained in Exhibit 1 Notes. Estimates might not sum to total enrollment or 100 percent because of rounding. No differences between all individual-market enrollees in 2013 and enrollees in off-marketplace plans in 2015 are significant (p < 0:05). All differences between enrollees in Marketplace and off-marketplace plans are significant (p < 0:05), except for age distribution and share with income of percent of the federal poverty level. a 7,735,000 enrollees. b 14,852,000 enrollees. c 6,286,000 enrollees. d 8,566,000 enrollees. August :8 Health Affairs 1491

4 Exhibit 4 Age distribution of individual-market enrollees, 2013 and 2015 SOURCE Authors analysis of data for 2013 and 2015 from the National Health Interview Survey. NOTES The individual market is explained in Exhibit 1 Notes. For enrollees in Marketplace plans, the share of those ages is significantly different from the share in that age group in off-marketplace plans (p < 0:05). No other differences between enrollees in Marketplace and off-marketplace plans, and no differences between 2013 individual-market enrollees and 2015 enrollees in off-marketplace plans, are significant (p < 0:05). Exhibit 5 deal between enrollees in Marketplace plans and those in off-marketplace plans (Exhibit 3). Enrollment gains were particularly large in the group with incomes of percent of poverty, especially in the Marketplace where this group was eligible for generous subsidies in the form of premium tax credits and cost-sharing Health status distribution of individual-market enrollees ages 19 64, 2013 and 2015 SOURCE Authors analysis of data for 2013 and 2015 from the National Health Interview Survey. NOTES The individual market is explained in Exhibit 1 Notes. Excellent/very good, good, and fair/poor refer to general health status. Smokers are people who smoke every/some day. Obese is body mass index greater than 30 kg/m 2. Hypertension and diabetes are ever diagnosed with the condition. All differences between individual-market enrollees in 2013 and 2015 are significant (p < 0:05), except for smoking. No differences between 2013 individual-market enrollees and 2015 enrollees in off-marketplace plans are significant (p < 0:05), except for smoking. All differences between enrollees in Marketplace and off-marketplace plans are significant (p < 0:05), except for obesity. reductions. Fifteen percent of off-marketplace enrollees were also in this income category, however, which is notable given their likely opportunities for premium and cost-sharing assistance had they purchased Marketplace plans. Enrollment also expanded for all racial and ethnic groups, with growth in total enrollment largest for non-hispanic whites, but with larger relative increases for the other groups especially in the Marketplace. The similarity between the characteristics of people purchasing in the individual market in 2013 and those purchasing coverage outside the Marketplace in 2015 was striking. There were no significant differences (p <0:05) in demographic or socioeconomic characteristics across these two groups. In contrast, off-marketplace enrollees in 2015 were more likely to be male, have higher incomes, be white, and have higher levels of education, compared to Marketplace enrollees (Exhibit 3). They were also more likely to be US citizens, work full time, and be married. All differences between 2015 enrollees in Marketplace plans and those in off-marketplace plans were significant (p <0:05), with the notable exception of the age distribution. However, off-marketplace enrollees were more likely than Marketplace enrollees to be in the youngest age group (ages 19 25). More than 50 percent of enrollees in both market segments were ages 45 and older (Exhibits 3 and 4). The health status of individual-market enrollees declined from 2013 to 2015, and in 2015 off- Marketplace purchasers were generally in better self-reported health than their Marketplace counterparts (Exhibit 5). Overall, the health status of off-marketplace purchasers in 2015 was similar to that of all enrollees in the individual market in In 2014, 46 percent of Marketplace enrollees had been uninsured at some point in the past twelve months, compared to only 8 percent of off-marketplace enrollees (Exhibit 6). By 2015, the share of Marketplace enrollees who had been uninsured had fallen to 22 percent, while the share of off-marketplace enrollees remained essentially unchanged. The share of respondents who said that their coverage was better than it had been one year ago was also similar across years for those purchasing off-marketplace plans. For those who purchased Marketplace plans, however, 45 percent said in 2014 that their coverage had improved, and 26 percent reported better coverage in Discussion Before the ACA s insurance reforms took effect, coverage in the individual market was often un Health Affairs August :8

5 Exhibit 6 Continuity and quality of coverage among individual-market enrollees ages 19 64, SOURCE Authors analysis of data for from the National Health Interview Survey. NOTES The individual market is explained in Exhibit 1 Notes. All differences between enrollees in Marketplace plans and those in off-marketplace plans, and between enrollees in Marketplace plans and all individual-market enrollees in 2013, are significant (p < 0:05). No differences between enrollees in off- Marketplace plans and all individual-market enrollees in 2013 are significant (p < 0:05), except for health insurance better than last year for enrollees in off-marketplace plans in available or unaffordable for people in poor health. The ACA guaranteed access to the individual market to all people regardless of health status beginning in 2014, while eliminating differences in individual-market premiums based on health status (other than for tobacco use) and restricting differences based on age. Subsidies in the form of federal premium and costsharing assistance were also available to eligible individuals purchasing Marketplace plans, but not for purchasers of off-marketplace plans. Following these changes, we found that individual-market enrollment among nonelderly adults nearly doubled between 2013 and While some people who purchased Marketplace plans would likely have had coverage in the absence of the ACA, our findings suggest that the Marketplaces and the associated financial assistance expanded coverage to large numbers of low-income nonelderly adults who would otherwise have been uninsured. Demand for individual coverage outside the Marketplaces remained strong, however, especially among adults with higher incomes and fewer health problems than those purchasing Marketplace plans. Some off-marketplace plans were grandfathered plans or transitional plans that did not have to comply with ACA regulations. Rising premiums and insurer exits from the Marketplaces have raised concerns about the stability of the Marketplaces and exposed a fundamental tension between pricing based on individual risk and requiring healthy and sick consumers to pay similar premiums. Under the ACA, reforms to the individual market moved toward broader risk pooling, with subsidies and a mandate to encourage participation. But healthy adults not eligible for subsidies, like many of the purchasers of off-marketplace plans in our study, may be at risk of dropping coverage as premiums rise and as grandfathered and transitional plans disappear. Proposals to replace the ACA recently considered by Congress aim to make the individual market more attractive for young and healthy people, by relaxing age rating restrictions and allowing states more flexibility on covered benefits. In doing so, they also would reduce risk pooling, raising premiums and reducing comprehensiveness of coverage for older, sicker, and lower-income people. To sustain recent gains in coverage or expand on them, efforts to price coverage based on risk and thereby attract healthy enrollees must be coupled with subsidies for people who would otherwise face high premiums and cost-sharing obligations relative to income. 11 This work was funded in part by the Robert Wood Johnson Foundation. The authors are grateful to Jason Gates for research assistance, and to Linda Blumberg and John Holahan for helpful comments. The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Urban Institute, its trustees, or its funders. August :8 Health Affairs 1493

6 NOTES 1 We collectively refer to the statebased exchanges and the federally managed exchange as Marketplaces. See HealthCare.gov [home page on the Internet]. Baltimore (MD): Centers for Medicare and Medicaid Services; [cited 2017 Jun 20]. Available from: 2 McMorrow S, Kenney GM, Long SK, Gates JA. Marketplaces helped drive coverage gains in 2015; affordability problems remained. Health Aff (Millwood). 2016;35(10): Blavin F, Karpman M, Zuckerman S. Understanding characteristics of likely Marketplace enrollees and how they choose plans. Health Aff (Millwood). 2016;35(3): Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation. Health insurance Marketplaces 2015 open enrollment period: March enrollment report [Internet]. Washington (DC): ASPE; 2015 Mar 10 [cited 2017 Jun 20]. Available from: files/pdf/83656/ib_2015mar_ enrollment.pdf 5 We used public use National Health Interview Survey data from IPUMS Health Surveys [home page on the Internet]. Minneapolis (MN): University of Minnesota; [cited 2017 Jun 20]. Available from: 6 State Health Access Data Assistance Center. Defining family for studies of health insurance coverage [Internet]. Minneapolis (MN): University of Minnesota; 2012 Mar [cited 2017 Jun 20]. (Issue Brief No. 27). Available from: default/files/publications/ SHADAC_Brief27.pdf#overlaycontext=publications/definingfamily-studies-health-insurancecoverage 7 National Center for Health Statistics National Health Interview Survey (NHIS): rules for evaluating and assigning exchange-based coverage [Internet]. Hyattsville (MD): NCHS; [cited 2017 Jun 20]. Available from: data/nhis/health_insurance/2014- exchange_coding_-rules.pdf 8 National Center for Health Statistics. Evaluation and editing of health insurance data [Internet]. Hyattsville (MD): NCHS; [last updated 2014 Aug 15; cited 2017 Jun 20]. Available from: nhis/health_insurance/hi_eval.htm 9 We used t-tests in the case of income because the chi-square test was not compatible with multiple imputation procedures used to deal with missing income values. 10 To access the Appendix, click on the Appendix link in the box to the right of the article online. 11 Holahan J, Blumberg LJ. Instead of ACA repeal and replace, fix it [Internet]. Washington (DC): Urban Institute; 2017 Jan [cited 2017 Jun 20]. Available from: publication/87076/ repealand-replace-aca-fix-it_2.pdf 1494 Health Affairs August :8