By Ezra Golberstein, Gilbert Gonzales, and Benjamin D. Sommers. adults are a major part of the Affordable Care

Transcription

1 doi: /hlthaff HEALTH AFFAIRS 34, NO. 10 (2015): Project HOPE The People-to-People Health Foundation, Inc. By Ezra Golberstein, Gilbert Gonzales, and Benjamin D. Sommers California s Early ACA Expansion Increased Coverage And Reduced Out-Of-Pocket Spending For The State s Low-Income Population Ezra Golberstein (egolber@ umn.edu) is an assistant professor in the Division of Health Policy and Management at the University of Minnesota School of Public Health, in Minneapolis. Gilbert Gonzales is an assistant professor in the Department of Health Policy at the Vanderbilt University School of Medicine, in Nashville, Tennessee. While workingonthisarticlehewas a doctoral candidate in the Division of Health Policy and Management at the University of Minnesota School of Public Health. ABSTRACT The Affordable Care Act (ACA) expanded eligibility for Medicaid to millions of low-income adults. While many expanding states implemented their expansion in 2014, five states and the District of Columbia expanded eligibility as early as 2010 by taking advantage of provisions in the ACA and Medicaid waivers. We used restricted data from the National Health Interview Survey to examine the impact of California s Low Income Health Program, an early expansion program that began in Our study demonstrates that the county-by-county rollout of expanded public insurance coverage in California significantly increased coverage, by 7 percentage points, and significantly reduced the likelihood of any family out-of-pocket medical spending in the previous year, by 10 percentage points, among low-income adults. Benjamin D. Sommers is an assistant professor in the Department of Health Policy and Management, Harvard School of Public Health, in Boston, Massachusetts. Medicaid expansions to lowincome adults are a major part of the Affordable Care Act s (ACA s) approach to expanding health insurance. The ACA allows states to expand Medicaid coverage to all adults with a family income of up to 138 percent of the federal poverty level, which is a significant increase from pre-aca Medicaid eligibility thresholds in most states particularly for childless adults. 1 Medicaid expansions went into effect on January 1, 2014, in about half of the states and the District of Columbia. However, states had the option of expanding their Medicaid programs under the ACA as early as April 1, Five states (California, Connecticut, Minnesota, New Jersey, and Washington) and the District of Columbia had opted to expand coverage to low-income adults before 2014, using a combination of the ACA s early expansion option and Medicaid section 1115 waivers. 2 In several of these states, the expansion was relatively modest in terms of income eligibility for childless adults (New Jersey expanded eligibility to those with incomes of 23 percent of poverty, and Connecticut expanded it to those with incomes of 56 percent of poverty; previously, neither state offered Medicaid eligibility to childless adults). In Minnesota and the District of Columbia, the expansion largely shifted people from existing public insurance programs into Medicaid. Early Expansion In California California s expansion, titled the Low Income Health Plan, changed existing programs and extended eligibility to a potentially larger group of low-income adults than was the case in the other states with early expansion. The Low Income Health Plan expanded eligibility to people with incomes as high as 200 percent of poverty, depending on their county of residence. 3 Before this insurance expansion, parents of Medicaideligible children with incomes of up to 100 percent of poverty were already eligible for Medicaid in California. California s early expansion was implemented at the county level under a section 1115 demonstration waiver, and participating counties received funds from the federal government equal to 50 percent of their expansion costs (which was 1688 Health Affairs October :10

2 California s traditional Medicaid match rate). 3 Beginning in November 2010 individual counties were allowed to expand coverage under the ACA and increase their income thresholds for eligibility up to a maximum of 200 percent of poverty. For example, Santa Clara, Los Angeles, and Orange Counties expanded eligibility in July 2011 to up to 75 percent of poverty, 133 percent of poverty, and 200 percent of poverty, respectively. Some of the most populous counties in California were among the first to expand. However, all but six of the state s fifty-eight counties eventually adopted the expansion at various times between July 2011 and March Online Appendix Exhibit A1 illustrates the timing of the counties expansions and their income thresholds for eligibility. 4 The Low Income Health Plan ended in January 2014, when participants were moved into the state s regular Medicaid program (Medi-Cal) or entered the private health insurance Marketplace. 3 This article evaluates the effects of the ACA s early public insurance expansion in California on health insurance coverage, access to care, and out-of-pocket spending. Before the implementation of the ACA, states had several options for expanding their Medicaid programs, such as increasing income thresholds for low-income parents or pursuing federal waivers that enabled states to cover childless adults. In the past two decades many states took advantage of one or more of these options, and extensive research has documented the effects of these expansions on the actual take-up of Medicaid coverage (or Medicaid-like coverage), the corresponding crowd-out of private insurance, access to care, changes in use of health care, and health outcomes. 5 8 In terms of the early Medicaid expansions under the ACA, one previous study used American Community Survey data to examine Connecticut and the District of Columbia. The researchers found increased Medicaid coverage in those populations, particularly among adults with health-related limitations. 9 We focused on California because it was the largest of the early expansion states and because its staggered expansion across counties offered a useful quasi-experimental research design for studying the effects of the coverage expansion on low-income adults. This study builds on previous research 10,11 that demonstrated the impact of the ACA s Medicaid expansions on health insurance coverage, and it extends the analysis to evaluate the effects of the ACA on access to care and out-of-pocket health care spending. Study Data And Methods Data We used survey data for from the National Health Interview Survey (NHIS). The survey, conducted annually by the National Center for Health Statistics, is a large, cross-sectional survey that is representative of the civilian noninstitutionalized US population. It is one of the main data sources used for tracking health and health care issues in the United States. Our analysis used the Integrated Health Interview Series, a version of the NHIS that makes variables and documentation consistent across years. 12 We linked the NHIS data to county-level information that indicated the timing of California s Low Income Health Plan expansion. To do this, we used a restricted-access version of the NHIS with information on the quarter of the year when the interview was conducted and the respondent s state and county of residence, via a data use agreement with the National Center for Health Statistics. Outcomes We examined three outcome domains: insurance coverage, access to and use of health care, and out-of-pocket medical spending. We first estimated the impact of California s expansion on three insurance-related outcomes, based on the respondent s coverage at the time of the interview: whether individuals reported having Medicaid or other state-sponsored public health insurance coverage (which would include the Low Income Health Plan), private insurance, and any health insurance. We then looked at three outcomes related to utilization and access: whether the respondent reported having any doctor s office visit in the previous two weeks, needing but not getting care in the previous twelve months because of cost, and any delay in receiving care in the previous twelve months because of cost. We next analyzed two outcomes related to outof-pocket medical spending: whether the respondent reported that he or she or a family member had spent any money on medical care in the previous twelve months, and spending over $500 on medical care in the same. These outcomes were derived from a single NHIS question that asked respondents how much they and their family spent on medical care in the past twelve months, exclusive of money spent on insurance premiums, over-the-counter drugs, and costs that will be reimbursed. Respondents chose one of the following categories: no spending and spending of $1 $499, $500 $1,999, $2,000 $2,999, $3,000 $4,999, or $5,000 or more. We chose to analyze the thresholds of any spending and spending more than $500 for two reasons. First, relatively few households incur higher levels of out-of-pocket spending, October :10 Health Affairs 1689

3 which limits the power to detect significant effects at higher thresholds. And second, for the low-income population we studied, even moderate out-of-pocket spending can represent a sizable financial burden. Analysis To assess the effects of California s insurance expansion, we estimated multivariate regression models that took advantage of the staggered timing of the insurance expansion across counties in California. Our quasi-experimental approach measured the effect of the expansion by comparing changes in each outcome after each county implemented the Low Income Health Plan to changes in counties that did not participate in the expansion, a procedure known as a difference-in-differences analysis. We estimated multivariate regression models in which there were two key predictor variables. First, for the outcomes of current insurance status and doctor visit in the previous two weeks, an indicator variable, interim expansion, was coded as 1 for observations from a county during the quarter in which the expansion began. For outcomes that corresponded to the previous twelve months, interim expansion was also coded as 1 if the expansion in a respondent s county happened within the previous twelve months. The second key variable was postexpansion, an indicator variable for whether the Low Income Health Plan had been expanded in a county for the entire quarter for the outcomes of current insurance status and doctor visit in the previous two weeks; or whether the Low Income Health Plan had been expanded in a county for the entire year for all other outcomes. The coefficient on this variable represents the effect of the expansion on the outcome variable. The regression models were adjusted for the following individual- and family-level sociodemographic characteristics: age, sex, race/ethnicity, US citizenship and country of origin, educational attainment, marital status, whether anyone in the family has a health-related disability, number of children in the household, and whether family income is below poverty. All models also included a full set of county and quarter fixed effects, which captured the statewide underlying trends in the outcomes both before and after counties started their expansions in 2011 (in both expansion and nonexpansion counties) and the direct effect of being in an expansion county. We report the coefficients on the two key variables interim expansion and postexpansion which are interpreted as the average change in the outcome after the expansion, after accounting for the preexisting trends in the outcome and time-invariant differences across counties. Appendix Exhibit A2 describes the full regression equation. 4 We restricted the analytic sample to adults ages who lived in California at the time of the interview and did not report receiving Supplemental Security Income in the previous year. Supplemental Security Income is an independent pathway to Medicaid eligibility; sensitivity analyses that included this group produced similar findings to those of our main analyses. We ran two sets of analyses. The first restricted the sample to people in households with incomes under 200 percent of poverty, since this was the upper income limit of the population potentially affected by California s Low Income Health Plan expansion. Our second analysis restricted the sample to adults in households with incomes under 138 percent of poverty, which is the population currently eligible for the Medicaid expansions in what is now thirty states and the District of Columbia. Analyses were conducted using Stata, version 13. All descriptive statistics and regression estimates were weighted by the NHIS sampling weights to reflect the survey s complex sampling design. We estimated linear regression models instead of nonlinear models, to facilitate the interpretation of the key coefficients. Robust standard errors clustered on the county were used to adjust for correlated outcomes within counties and serial correlation. 13 We relied on the five imputation files created and recommended by the National Center for Health Statistics to impute family income when data were missing. 14 All results presented account for these imputations using Stata s multiple imputation package. Limitations Our study had some limitations. First, we conducted a quasi-experimental analysis of the coverage expansion decisions in fiftytwo California counties. Our estimates might be biased to the extent that other things were changing in those counties around the time that the Low Income Health Plan was implemented. Second, our ability to detect small changes in health care outcomes was constrained by the sample sizes representing California in the NHIS. As with all survey-based studies, a larger sample size would have offered more precise estimates. Yet unlike other surveys used to monitor health insurance coverage, the NHIS has the advantage of including detailed measures of access to care and health services utilization. Finally, our ability to properly identify the population eligible for subsidized coverage is only as accurate as the self-reported measures of income. Although detailed, these might include some errors, and they are also likely to differ somewhat from the precise definition used for program eligibility purposes Health Affairs October :10

4 Study Results For most variables, the samples with incomes of less than 200 percent of poverty and less than 138 percent of poverty were quite similar (Exhibit 1). Twenty-three percent of the first sample and 29 percent of the second sample had Medicaid or other state-sponsored health insurance at the time of the interview, while just over half of both samples had some type of insurance at the time of the interview. In both sample groups, approximately 10 percent had a doctor s visit in the previous two weeks, 13 percent did not get needed care in the previous twelve months because of cost, and 15 percent delayed getting care in the previous twelve months because of cost (Exhibit 1). In the group with incomes of less than 200 percent of poverty, 77 percent had any out-of-pocket medical spending in the previous twelve months, and 37 percent had at least $500 of out-of-pocket spending in the previous twelve months. Both of those proportions were slightly lower for the group with incomes of less than 138 percent of poverty. We found that after counties implemented the Low Income Health Plan, the likelihood of people in those counties having Medicaid or other state-sponsored health insurance, compared to people in nonexpansion counties, rose by 7.5 percentage points in the group with incomes below 200 percent of poverty and by 7.3 percentage points for the group with incomes below 138 percent of poverty (top panel of Exhibit 2). We also found that in the second group, this increase in public coverage happened immediately after eligibility was expanded, as indicated by the increase during the implementation (or interim expansion). We did not find evidence of any significant changes in private insurance for either income group. Overall rates of having insurance rose by 5.8 percentage points for the first group and by 7.2 percentage points for the second group. Overall, our results related to doctor visits and unmet need for care or delayed care due to cost all point toward improved access and utilization after the implementation of the Low Income Health Plan (middle panel of Exhibit 2). However, none of our findings was significant for either of the two income groups. We did find reductions in out-of-pocket spending associated with the coverage expansion (bottom panel of Exhibit 2). For the group with incomes below 200 percent of poverty, after counties expanded coverage, the likelihood of reporting any out-of-pocket health care spending dropped by 10 percentage points, and this decline began during the implementation. For the outcome of spending at least $500 out of pocket, we found an approximately 4-percentage-point reduction in both the implementation and postimplementation s, but the reduction was only marginally significant, and only for the implementation. The overall pattern of results was similar for the group with incomes below 138 percent of poverty (bottom panel of Exhibit 2). The most notable difference was that this group showed a much stronger and more precise effect of the Low Income Health Plan on having at least $500 in out-of-pocket spending in the of the plan s implementation in the respondent s county (an 8.2-percentage-point reduction). To test our results for robustness, we excluded Exhibit 1 Characteristics Of California s Low-Income Population Ages 19 64, By Family Income Level, Family income Characteristic <200% of poverty <138% of poverty Sample size 23,857 15,841 Age (years) % 23.3% Sex Female 50.9% 51.3% Race/ethnicity White, non-hispanic 23.4% 21.3% Black, non-hispanic Other, non-hispanic Hispanic Education Less than high school 36.2% 40.0% High school Some college or technical training College graduate Income Income below federal poverty level 40.9% 62.9% Family income (mean) $25,015 $18,599 Insurance Medicaid or other state-sponsored health insurance 23.0% 28.7% Any insurance Health care use and spending Any doctor visits, past 2 weeks 9.9% 10.4% Didn t get needed care because of cost, past 12 months Delayed care due to cost, past 12 months Any out-of-pocket spending, past 12 months Out-of-pocket spending >$500, past 12 months SOURCE Authors analysis of data for from the National Health Interview Survey. October :10 Health Affairs 1691

5 Exhibit 2 Changes In Health Insurance, Access To Care, And Out-Of-Pocket Health Care Spending After California s Early Medicaid Expansion For The Low-Income Population Ages 19 64, By Family Income Level Family income <200% of poverty <138% of poverty Interim expansion Postexpansion Interim expansion Postexpansion Insurance a Medicaid or other public Insurance *** 0.076* 0.073** Private insurance Any insurance ** ** Access to care Any doctor visits, past two weeks b Didn t get needed care because of cost, past 12 months c Delayed care due to cost, past 12 months d Out-of-pocket spending e Any out-of-pocket spending, past 12 months 0.061* 0.101** * Out-of-pocket spending >$500, past 12 months 0.043* *** SOURCE Authors analysis of data for from the National Health Interview Survey. NOTES The exhibit shows results of linear regression models. All models controlled for age, sex, race and ethnicity, citizenship and country of origin, education, marital status, the presence of someone in the family with ahealth-related disability, the number of children in the household, poverty status, county, and quarter of interview. Interim expansion and postexpansion are explained in the text. a For <200% of poverty, n ¼ 23; 462. For <138% of poverty, n ¼ 15; 571. b For <200 % of poverty, n ¼ 23; 844. For <138% of poverty, n ¼ 15; 835. c For <200% of poverty, n ¼ 23; 839. For <138 percent of poverty, n ¼ 15; 831. d For <200% of poverty, n ¼ 23; 827. For <138 % of poverty, n ¼ 15; 824. e For <200% of poverty, n ¼ 23; 027. For <138% of poverty, n ¼ 15; 303. *p < 0:10 **p < 0:05 ***p < 0:01 noncitizens who had been in the United States for less than five years, which reduced our original sample size by 6.5 percent. Although undocumented immigrants are not eligible for Medicaid, qualified noncitizen immigrants with documented legal status are eligible after five years of residence in the United States. The results from these models (not shown) were nearly identical to our main results, except that the increases in having public insurance and any insurance were slightly larger.we also found that controlling for employment and self-reported health status did not affect our results in any meaningful way. Discussion We present new evidence on the impact of the ACA s early public insurance expansion in the nation s most populous state. Our study is the first to assess the effects of this expansion on access to and use of health care and on out-ofpocket spending. Rates of having both public insurance and any insurance rose significantly after counties in California expanded coverage via the Low Income Health Program. We were unable to detect any significant changes in access to care or physician office visits. However, we did find reductions in the likelihood of reporting any out-of-pocket spending and the likelihood of incurring $500 or more in that spending in the previous year. Our results support the notion that the ACA s early Medicaid and Medicaid-like expansions (such as the Low Income Health Plan) succeeded in significantly increasing coverage among lowincome nonelderly adults. In 2010, according to our analysis, 46 percent of adults ages in California with incomes below 200 percent of poverty were uninsured. The 5.8-percentagepoint increase we found in coverage indicates that the early expansion reduced the uninsured rate in this population by over 12 percent. One question raised by our results was why we did not also find a significant improvement in access to care associated with the early expansion in California, since previous quasi-experimental studies and randomized trials found that Medicaid expansions were associated with improved access to care. 6,7,15 There are several possible explanations. First, in contrast to the Oregon Medicaid study, 15 which investigated the effect of individuals directly taking up Medicaid coverage, we looked at the county-level effects of county coverage expansions. This makes it more difficult to detect the effects of the expansion, since only a subset of eligible individuals took up Medicaid and similar public coverage options. 16 Second, in contrast to other work that did find positive effects of state Medicaid expansions on access to care, 6 our study had a much smaller 1692 Health Affairs October :10

6 sample size, and our nonsignificant results may reflect insufficient statistical power. However, a survey of nonelderly adult Californians in 2014 also found no significant differences in the likelihood of postponing or not receiving needed care between the uninsured and the newly insured. 17 Third, other low-income health care programs and well-funded safety-net providers existed in California before implementation of the Low Income Health Plan, which could mute the effects of the expansion on increasing access to care. 18,19 Finally, both our low-income samples (people with incomes below 200 percent of poverty and those with incomes below 138 percent of poverty) have limitations. Estimates for the first sample are likely conservative since the sample included many people who lived in counties that expanded eligibility up to thresholds substantially below 200 percent of poverty but whose incomes were too high to for them to become eligible. Use of the sample below 138 percent of poverty mostly avoids that problem, but its smaller sample size reduces its power and does not fully capture the effects of the expansion in counties that expanded eligibility to between the two income thresholds. An especially notable result is that California s early ACA expansion significantly reduced families out-of-pocket medical spending. One goal of any kind of insurance is to protect individuals from financial risk. In the case of health insurance, this risk takes the form of out-of-pocket spending on health care. We found large reductions in the likelihood that families had any outof-pocket medical spending in the previous year, and some evidence that high out-of-pocket spending decreased immediately following California s county-level coverage expansions. For low-income families, reducing out-of-pocket spending can be a substantial benefit, as the average household income was $25,015 for the group with incomes below 200 percent of poverty and $18,599 for the group with incomes below 138 percent of poverty (Exhibit 1). This result echoes other research on the ACA s expansion of dependent coverage insurance benefits to young adults, which also found significant reductions in high out-of-pocket spending. 20,21 Furthermore, it suggests that the ACA is providing significant benefits in terms of financial risk protection. Policy Implications Our results suggest that the insurance expansions under the ACA may have important effects on adults with incomes below 138 percent of poverty the population now eligible for Medicaid in thirty states and the District of Columbia. (On April 29, 2015, the Montana legislature also passed a Medicaid expansion program; as of August 19, 2015, the program had not been approved for a federal waiver.) We found that uninsurance dropped 7 percentage points in California for the population with incomes below 138 percent of poverty and that the expansion significantly reduced out-of-pocket spending. Low-income adults meeting new income eligibility standards under the ACA may be healthier than many low-income adults previously insured by Medicaid. Even if that is the case, our findings indicate that the coverage expansions are likely to provide substantial financial protection for new enrollees. In turn, these reductions in outof-pocket spending may correspond to reductions in uncompensated care, which would improve the financial status of safety-net providers, including hospitals. California is a large and diverse state that in many ways reflects the diversity of the US population, but it is a unique policy environment. However, it is reasonable to expect that the changes we observed in coverage and financial protection for low-income adults after the Low Income Health Plan was implemented will be even greater in states expanding Medicaid since 2014 that had less generous public programs, compared to California before the ACA. These states include Indiana, Kentucky, and West Virginia. Conclusion We found that the ACA s early insurance expansion for low-income families in California was associated with significant increases in having health insurance and with reductions in out-ofpocket health care spending. It is important to continue to assess how the ACA s Medicaid expansions that began in January 2014 are affecting health insurance rates and both access to and use of health care, and to identify the barriers to care that remain for low-income beneficiaries. October :10 Health Affairs 1693

7 Ezra Golberstein acknowledges support from the Minnesota Population Center s grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (Grant No. 5R24HD041023). Gilbert Gonzales acknowledges funding support from a Doctoral Dissertation Fellowship from the University of Minnesota Graduate School. Benjamin Sommers s workon this project was supported by the Agency for Healthcare Research and Quality (Grant No. K02HS021291). Sommers serves part time as an adviser in the Office of the Assistant Secretary for Planning and Evaluation, Department of Health and Human Services. The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Research Data Center, the National Center for Health Statistics, the Centers for Disease Control and Prevention, the Agency for Healthcare Research and Quality, or the Department of Health and Human Services. NOTES 1 Kaiser Commission on Medicaid and the Uninsured. A closer look at the impact of state decisions not to expand Medicaid on coverage for uninsured adults [Internet]. Washington (DC): The Commission; 2014 Apr 24 [cited 2015 Aug 17]. Available from: 2 Sommers BD, Arntson E, Kenney GM, Epstein AM. Lessons from early Medicaid expansions under health reform: interviews with Medicaid officials. Medicare Medicaid Res Rev. 2013;3(4). 3 Meng Y-Y, Cabezas L, Roby DH, Pourat N, Kominski GF. Successful strategies for increasing enrollment in California s Low Income Health Program (LIHP) [Internet]. Los Angeles (CA): UCLA Center for Health Policy Research; 2012 Sep [cited 2015 Aug 17]. Available from: healthpolicy.ucla.edu/publications/ Documents/PDF/lihppolicynote sep2012.pdf 4 To access the Appendix, click on the Appendix link in the box to the right of the article online. 5 Gruber J, Simon K. Crowd-out 10 years later: have recent public insurance expansions crowded out private health insurance? J Health Econ. 2008;27(2): Sommers BD, Baicker K, Epstein AM. Mortality and access to care among adults after state Medicaid expansions. N Engl J Med. 2012; 367(11): DeLeire T, Dague L, Leininger L, Voskuil K, Friedsam D. Wisconsin experience indicates that expanding public insurance to low-income childless adults has health care impacts. Health Aff (Millwood). 2013; 32(6): Leininger LJ, Friedsam D, Dague L, Mok S, Hynes E, Bergum A, et al. Wisconsin s BadgerCare Plus reform: impact on low-income families enrollment and retention in public coverage. Health Serv Res. 2011; 46(1 Pt 2): Sommers BD, Kenney GM, Epstein AM. New evidence on the Affordable Care Act: coverage impacts of early Medicaid expansions. Health Aff (Millwood). 2014;33(1): Long SK, Karpman M, Shartzer A, Wissoker D, Kenney GM, Zuckerman S, et al. Taking stock: health insurance coverage under the ACA as of September 2014 [Internet]. Washington (DC): Urban Institute, Health Policy Center; 2014 Dec 3 [cited 2015 Aug 17]. Available from: health-insurance-coverage-underthe-aca-as-of-september-2014.html 11 Sommers BD, Musco T, Finegold K, Gunja MZ, Burke A, McDowell AM. Health reform and changes in health insurance coverage in N Engl J Med. 2014;371(9): Minnesota Population Center, State Health Access Data Assistance Center. Integrated Health Interview Series: version 5.0 [Internet]. Minneapolis (MN): University of Minnesota; [cited 2015 Aug 17]. Available from: 13 Bertrand M, Duflo E, Mullainathan S. How much should we trust differences-in-differences estimates? Q J Econ. 2004;119(1): National Center for Health Statistics. Multiple imputation of family income and personal earnings in the National Health Interview Survey: methods and examples [Internet]. Hyattsville (MD): NCHS; 2014 Aug [cited 2015 Aug 28]. Available from: nhis/tecdoc13.pdf 15 Baicker K, Taubman SL, Allen HL, Bernstein M, Gruber JH, Newhouse JP, et al. The Oregon experiment effects of Medicaid on clinical outcomes. N Engl J Med. 2013;368(18): Sommers BD, Tomasi MR, Swartz K, Epstein AM. Reasons for the wide variation in Medicaid participation rates among states hold lessons for coverage expansion in Health Aff (Millwood). 2012;31(5): Garfield R, Majerol M, Young K. Coverage expansions and the remaining uninsured: a look at California during year one of ACA implementation [Internet]. Menlo Park (CA): Henry J. Kaiser Family Foundation; 2015 May [cited 2015 Aug 17]. Available from: files.kff.org/attachment/reportcoverage-expansions-and-theremaining-uninsured-a-look-atcalifornia-during-year-one-of-acaimplementation 18 Takach M, Osius E (National Academy for State Health Policy, Washington, DC). Federally qualified health centers and state health policy: a primer for California [Internet]. Oakland (CA): California HealthCare Foundation; 2009 Jul [cited 2015 Aug 17]. (Issue Brief). Available from: LIBRARY%20Files/PDF/F/PDF %20FederallyQualifiedHealth CentersAndStatePolicy.pdf 19 Kominski G, Pourat N, Roby D, Diamant A, Meng Y-Y, Kally Z, et al. Interim evaluation report on California s health care coverage initiative [Internet]. Los Angeles (CA): UCLA Center for Health Policy Research; 2010 Jun 1 [cited 2015 Aug 17]. Available from: Documents/Interim%20Evaluation %20Report%20on%20the%20 Health%20Care%20Coverage%20 Initiative%20in%20California.pdf 20 Busch SH, Golberstein E, Meara E. ACA dependent coverage provision reduced high out-of-pocket health care spending for young adults. Health Aff (Millwood). 2014;33(8): Chua K-P, Sommers BD. Changes in health and medical spending among young adults under health reform. JAMA. 2014;311(23): Health Affairs October :10