EARLY IMPLEMENTATION OF THE HEALTH COVERAGE TAX CREDIT IN MARYLAND, MICHIGAN, AND NORTH CAROLINA: A CASE STUDY SUMMARY Stan Dorn, Tanya Alteras, and Jack A. Meyer Economic and Social Research Institute April 2005 ABSTRACT: Health Coverage Tax Credits (HCTCs), which pay 65 percent of beneficiaries’ health insurance premiums, constitute an ambitious experiment in using the federal income tax system to subsidize health coverage for the uninsured. The HCTC program has made an excellent start, successfully developing program infrastructure and preventing the kind of marketing fraud that marred a previous tax credit program. However, the new program is experiencing low take- up rates; there have been delays and confusion surrounding enrollment into the advance-payment option; and there is some dissatisfaction with coverage offered by participating health plans. To gather more evidence about HCTCs’ effectiveness and assess their prospects as a model for broader reforms, researchers visited Maryland, Michigan, and North Carolina, which used varied approaches to HCTC implementation. The authors present key findings and propose reforms to improve HCTCs’ ability to help its current target population and aid policymakers in designing future health insurance tax credits. Support for this paper was provided by The Commonwealth Fund and The Nathan Cummings Foundation. The views presented here are those of the authors and should not be attributed to The Commonwealth Fund, The Nathan Cummings Foundation, or either philanthropy’s directors, officers, or staff, or to members of the Task Force on the Future of Health Insurance. Additional copies of this and other Commonwealth Fund publications are available online at www.cmwf.org. To learn about new Fund publications when they appear, visit the Fund’s Web site and register to receive e-mail alerts. Commonwealth Fund pub. no. 806. CONTENTS List of Tables and Figures................................................................................................ iv About the Authors........................................................................................................... v About the Economic and Social Research Institute .......................................................... v Acknowledgments .......................................................................................................... vi Executive Summary....................................................................................................... vii Introduction .................................................................................................................... 1 Study Background and Methodology ............................................................................... 2 Overview of Implementation in Maryland, Michigan, and North Carolina ...................... 3 Findings........................................................................................................................... 8 Infrastructure ............................................................................................................. 8 Enrollment............................................................................................................... 13 Health Plans............................................................................................................. 16 Areas for Potential Improvement ................................................................................... 23 Implications for Structuring Broader Coverage Expansions ............................................ 25 Conclusion .................................................................................................................... 26 Appendix A. National HCTC Rules.............................................................................. 27 Appendix B. Enrollment into Advance Payment ............................................................ 29 Notes............................................................................................................................. 31 iii LIST OF TABLES AND FIGURES Table ES-1 Health Coverage Tax Credit (HCTC) Enrollment for Selected States, September 2004 ....................................................... viii Table ES-2 Health Plan Premiums and Cost-Sharing for HCTC Enrollees, June 2004 .............................................................................. viii Table 1 Covered Benefits and Cost-Sharing in State-Qualified Plans, Case Study States, June 2004........................................................... 6 Table 2 Cost of State-Based Coverage, One-Person Policies, in Case Study States, June 2004 ................................................................ 7 Figure ES-1 Annualized 35 Percent Beneficiary Premium Share for Various Individuals Receiving Average State-Based, One-Person Coverage in North Carolina, June 2004............................... xi Figure 1 Advance-Payment Enrollment, Case Study States vs. Other States, September 2004................................................................... 3 Figure 2 Basis for HCTC Eligibility, Case Study States vs. Other States, September 2004................................................................... 4 Figure 3 Advance-Payment Enrollment into Types of Qualified Coverage, Case Study States vs. Other States with State-Qualified Plans, September 2004...................................................... 4 Figure 4 Annualized 35 Percent Beneficiary Premium Share for Various Individuals Receiving Average State-Based, One-Person Coverage in Maryland, June 2004 ....................................... 18 Figure 5 Annualized 35 Percent Beneficiary Premium Share for Various Individuals Receiving Average State-Based, One-Person Coverage in North Carolina, June 2004.............................. 19 iv ABOUT THE AUTHORS Stan Dorn, J.D., is a senior policy analyst at the Economic and Social Research Institute (ESRI). He has been involved in health policy at the state and national levels for almost 20 years, focusing on low-income consumers, Medicaid, the State Children’s Health Insurance Program (SCHIP), and the uninsured. Previously, Dorn served as director of the Health Consumer Alliance, a consortium of legal services groups in California that help low-income consumers obtain necessary health care. He also directed the Health Division of the Children’s Defense Fund, where he led the health policy team in CDF’s campaign that helped enact SCHIP in 1997. Before his work at CDF, Dorn directed the Washington, D.C., office of the National Health Law Program and served as a staff attorney in its Los Angeles headquarters. Tanya Alteras, M.P.P., is a senior policy analyst at ESRI. Ms. Alteras joined ESRI from the Department of Health and Human Services’ office of the Assistant Secretary for Planning and Evaluation, where she worked on a nationwide evaluation of SCHIP and reviewed federal and state policy reforms. At ESRI, Alteras is examining such issues as delivery system innovations designed to increase coverage and improve health care access and outcomes. Jack A. Meyer, Ph.D., is the founder and president of ESRI. Dr. Meyer directs health care research projects and conducts policy analysis on issues related to covering the uninsured and improving health care quality. He is the author of numerous books, monographs, and articles on policy reforms that can enhance access to health care, promote better patient outcomes, and improve the efficiency of the U.S. health care system. Meyer is also a Visiting Fellow at the Brookings Institution. ABOUT THE ECONOMIC AND SOCIAL RESEARCH INSTITUTE Founded in 1987, ESRI is a nonpartisan, nonprofit research organization headquartered in Washington, D.C. Specializing in health and social policy research, ESRI conducts studies aimed at enhancing the effectiveness of social programs, improving the ways in which health care services are organized and delivered, and making high-quality health care accessible and affordable. v ACKNOWLEDGMENTS The authors are grateful to The Commonwealth Fund and The Nathan Cummings Foundation, whose financial support made this study possible. The authors would also like to thank the following national officials who provided information and who commented on earlier drafts both of this overview report and our three state-specific reports: Stephen Finan and Roy Ramthun of the U.S. Department of the Treasury; Carolynn Adams, Karin Cano, Jane Looney, and Keith V. Taylor of the Internal Revenue Service; Julie R. Crom, Samantha Speede, and Tamara Taylor of Accenture; and JoAnn Lamphere of The Lewin Group. In addition, numerous individuals in Maryland, Michigan, and North Carolina lent considerable assistance to this study, providing information about their state’s implementation of the Health Coverage Tax Credit program and commenting on earlier report drafts: • Maryland: Susanne Albin and Richard Popper of the Maryland Health Insurance Program; Patrick Baker and Valerie Kurnas of the Maryland Department of Labor, Licensing and Regulation; and Robin Vahle of Care First Blue Cross and Blue Shield. • Michigan: Marc Barragan, United Steelworkers of America Local 1299; Chuck Batt, Diane Cygan, and Lisa Kirtzhals of Michigan Works! Downriver Community Conference; Sandra Damesworth, Linda Karos, Samuel Johnson, John Henige, and Deborah Hennessey of the Michigan Unemployment Insurance Agency; Vicki Enright, Joan Moiles, David Plawecki, Robert Swanson, and Sharon Walker of the Michigan Department of Labor and Economic Growth; and Alex Stamatoulakis of the Southeast Michigan Community Alliance. • North Carolina: Phil Telfer, the Office of the Governor; Myra Allen Beatty and Roger Shackleford, Commission on Workforce Development; Tim Beam, Martha Bowman, Linda Burton, David Canady, Dianne Creech, Lane Dyer, Manfred Emmrich, Curtis Morrow, Muriel Offerman, and Cecilia Williams, Employment Security Commission; Alan Alexander, Russell Doles, and Steve Gold, Division of Employment and Training; Jim Cook, Department of Social Services; Bernetha Brown, Union of Needletraders, Industrial and Textile Employees; Lynda Brower- Isabel, Jennifer Grady, Danny McKinney, King Prather, Art Slater, Susan Smith, and Mark Stinneford, Blue Cross and Blue Shield of North Carolina; and Annita Bennish and Ruth Crisco, members of the community. None of the funders or respondents, however, is responsible for the views expressed in this report, which are solely the authors’. vi EXECUTIVE SUMMARY When the Trade Act of 2002 became law, the nation began a major experiment in helping uninsured Americans purchase health coverage. This legislation created Health Coverage Tax Credits (HCTCs), which pay 65 percent of beneficiaries’ premiums for qualified coverage. Such coverage consists primarily of COBRA plans sponsored by former employers and private health plans offered to HCTC beneficiaries by state arrangement. The credits are refundable, which means they are paid in full to eligible households, including those who owe little or no federal income tax. The credits are also advanceable, which means that at the beneficiary’s request they can be paid directly to the insurer each month, as premiums are due. The HCTC program has made an excellent start establishing basic program infrastructure and preventing the kind of widespread fraud that was reported in connection with an earlier health insurance tax credit program. However, enrollment in the new tax credit has been low; consumers have experienced delays and confusion as a result of the complex enrollment process for advance payment; and there has been some dissatisfaction with the coverage offered by participating health plans. To gather evidence about the effectiveness of the new tax credits and to assess progress in finding solutions to emerging challenges, researchers from the Economic and Social Research Institute (ESRI) visited Maryland, Michigan, and North Carolina. In those states, enrollment of potentially eligible individuals into HCTC advance payment, though small in absolute terms (between 8 percent and 12 percent of potentially eligible individuals), was larger than the comparable national enrollment rate of 6.1 percent (representing 13,500 individuals of the nearly 222,000 who were potentially eligible for advance payment). Through these visits, the researchers sought to learn how three very different states, each using its own distinctive approach to implementation, achieved this comparative success. The authors also hoped that these visits would help identify obstacles that even the most effective state officials and private sector leaders face with the new program, as well as point the way toward possible solutions. Tables ES-1 and ES-2 on the next page show some of the salient characteristics of these three states, in terms of eligibility, enrollment, and the state-based plans that were offered to HCTC beneficiaries. vii Table ES-1. Health Coverage Tax Credit (HCTC) Enrollment for Selected States, September 2004 North Average of Maryland Michigan Carolina All Other States Percentage of potentially eligible workers enrolled in HCTC advance paymenta 11.5% 7.7% 9.5% 5.6%c Percentage of potentially eligible workers who qualified for HCTC based on 90% PBGC 61% PBGC 36% PBGC 69% PBGC payments from Pension Benefit Guaranty Corporation (PBGC) or receipt of 10% TAA 39% TAA 64% TAA 31% TAA Trade Adjustment Assistance (TAA)b Percentage of advance-payment enrollees who chose state-qualified plans 46% 62% 61% 41%d Table ES-2. Health Plan Premiums and Cost-Sharing for HCTC Enrollees, June 2004 Maryland’s High- Michigan’s Community- North Carolina’s Risk Poole Rated Coverage Nongroup Plans Lowest deductible among offered plans $0 $250 $250 Highest deductible (or other cost-sharing) among offered plans $1,000 50% co-insurance $5,000 Age, gender, health status, Basis for premium variation Age None area of residence Lowest premium offered to $811 $1,292 $239f Beneficiaries’ annualized, 35% premium 25- to 55-year-olds share for a one-person policy Highest premium offered to $2,557 $1,710 $5,847f 25- to 55-year-olds a Percentages of potentially eligible workers enrolled in advance payment were calculated by dividing the number of advance payment enrollees, per state, into the total number of state households whom PBGC or a State Workforce Agency (SWA) identified as potentially eligible for HCTC and who were mailed outreach materials. These enrollment percentages do not include households claiming HCTCs only on their annual tax returns. b TAA-based eligibles include those who received TAA cash payments, qualified for such payments but for receipt of unemployment insurance, or received Alternative Trade Adjustment Assistance income support provided to certain older, displaced workers. c Including the three case study states, the national percentage enrollment into advance payment was 6.1 percent of potentially eligible individuals (13,500 of 221,716 potentially eligible households). d The percentage of advance payment enrollees who chose state-qualified plans in all other states, as listed in this table, included only states in which at least one such plan (other than mini-COBRA coverage required of small employers under state law) was open to enrollees in September 2004. e Shortly before the site visit, Maryland also arranged to have a nongroup plan offered to HCTC beneficiaries. Only 7 HCTC beneficiaries were then enrolled in that plan, compared to roughly 500 beneficiaries in the high-risk pool. f North Carolina premium quotes were for individuals living in the state capital, Raleigh. The person with the lowest premium was a 25-year-old male in the healthiest risk tier buying coverage with a $5,000 deductible. The person with the highest premium was a 25-year-old woman in the least healthy risk tier who bought coverage that included maternity care and that had a $250 deductible. The second-highest beneficiary premium share was $4,725, for a 55-year-old woman in the least healthy risk tier who enrolled in the plan with a $250 deductible. viii Following are some key findings from the three-state study: Public and private entities involved in state-level HCTC implementation were dedicated, creative, and effective. Gubernatorial leadership was critical in making HCTC implementation a top priority, with both public and private actors showing remarkable commitment to the program. In all three case study states, government officials worked together across traditional organizational boundaries to get HCTC-related systems up and running in remarkably brief periods of time. All three states likewise devised creative mechanisms to prevent beneficiaries from being required to pay premiums in full for several months before the start of advance payment. These mechanisms included so-called “bridge” or “gap filler” programs in North Carolina and Maryland that used National Emergency Grants (NEGs) from the U.S. Department of Labor (DOL) to subsidize beneficiaries’ premiums while they waited for their advance payment to start. Agencies involved in national HCTC implementation received generally positive reviews. In all three states, public and private stakeholders commended the responsive and nimble assistance they received from the Treasury Department, the Internal Revenue Service (IRS), and the private contractors working with the IRS. The leadership at DOL was viewed with more ambivalence. When unexpected problems emerged with advance payment, DOL received plaudits for creatively adjusting its policy on states’ permissible use of NEG dollars. However, concerns were expressed about DOL delays in ruling on specific state requests for NEG funds. Enrollment into advance payment was complex, which increased administrative costs and reduced take-up of the tax credit. To apply for HCTC, a displaced worker was required to make various applications to at least three different entities (the state workforce agency, the health plan, and the national HCTC office). To further complicate the application process, workers were responsible for passing documentation, in hard-copy form, back and forth between various entities to which they were applying. Privacy concerns often prevented federal officials, state officials, and health plans from sharing data or even speaking directly with one another about particular beneficiaries’ applications. As a result, beneficiaries frequently had to act as “go betweens” communicating technical matters between multiple public and private agencies; these efforts increased administrative costs, and when beneficiaries could not act as satisfactory intermediaries, many lost coverage. To their credit, the officials administering the national HCTC program have taken several promising steps in recent months to simplify the enrollment process, though more may need to be done. ix Most informants agreed that the main reason why few beneficiaries enrolled in advance payment was workers’ perception that they could not afford to pay 35 percent of premiums. According to the vast majority of informants in the three study states, beneficiaries’ inability to pay their 35 percent premium share was by far the most important factor limiting enrollment. That problem was compounded by other obstacles to take-up, including delays in the start of individual beneficiaries’ advance payment and the general complexity and confusion of enrollment. Most enrollees preferred comprehensive coverage to less expensive insurance with high cost-sharing. In Maryland, 60 percent of beneficiaries enrolled in the high-risk pool’s more comprehensive HMO option, which had no deductible, rather than the PPO, which had a $1,000 deductible, even though the HMO cost roughly 50 percent more. In Michigan, every beneficiary who enrolled in state-qualified coverage chose the more comprehensive plan, even though it cost 32 percent more than the plan with 50 percent coinsurance and no coverage of physician visits or preventive care. In North Carolina, 52 percent selected coverage with $250 or $500 deductibles, even though such coverage cost between 22 and 70 percent more than the highest-deductible plans. By contrast, a number of North Carolinians who selected higher-deductible plans to lower their premium costs later expressed strong dissatisfaction with that coverage. For example, some adults with chronic illnesses could not afford to refill their prescriptions, perceived their coverage as largely pointless, and eventually disenrolled. Health coverage with pure or modified community rating avoided some serious problems with nongroup coverage that based premium levels on individualized assessment of health status. In Michigan and Maryland, stakeholders expressed no dissatisfaction with those states’ rating rules for HCTC coverage. Michigan’s coverage was purely community rated, with all enrollees into a given plan charged the same premiums, regardless of age, gender, area of residence, or individual health history. Similarly, Maryland’s high-risk pool varied premiums based on age alone, without any medical underwriting (the process through which insurers assess each applicant’s individual health risk). On the other hand, North Carolina’s rating rules permitted premium variation based on factors that included age, gender, and medical underwriting. These variations created tremendous dissatisfaction, according to virtually every stakeholder interviewed. Beneficiaries, officials, and others typically found it unfair that HCTC’s state-qualified coverage was least affordable to those who needed it most, with premiums that varied dramatically based on factors outside the individual’s control. For example, premiums x could increase by 45 percent because of gender (for plans that excluded routine maternity care); by 179 percent for women who opted to include coverage for routine maternity care; by 211 percent because of age; by 253 percent because of individual health history and health status; and by more than 1,300 percent because of all these factors combined. Figure ES-1. Annualized 35 Percent Beneficiary Premium Share for Various Individuals Receiving Average State- Based, One-Person Coverage in North Carolina, June 2004 Premium share Lowest risk Highest risk $6,000 $5,077 $5,000 $3,926 $4,066 $4,000 $3,000 $1,822 $2,000 $1,259 $1,439 $1,112 $1,152 $1,000 $357 $516 $0 25-Year-Old Man 25-Year-Old 25-Year-Old 55-Year-Old Man 55-Year-Old Woman Woman, with Woman, with Maternity Care Maternity Care Included Excluded Notes: (1) This figure shows the premiums that would be charged to various individuals for a single, weighted- average plan, determined based on total HCTC enrollment into all state-qualified options. (2) Lowest-risk and highest-risk premiums were based on risk tier ratios provided by BCBSNC and published quotes for individuals living in the state capital (Raleigh). Source: BCBSNC. Calculations by ESRI, March 2005. Medical underwriting in North Carolina also had a dramatic effect on take-up. Among the displaced workers who were quoted higher premium rates after the underwriting process concluded, fully 69 percent dropped out of the program at that point. If these individuals had instead completed their enrollment into advance payment, more than 3,900 additional North Carolinians would have received coverage, increasing total, national HCTC enrollment by 42 percent. Some adverse selection and risk segmentation among plan options may have taken place. In Maryland’s high-risk pool, there appeared to be a stark difference between HCTC enrollees into the pool’s more generous HMO plan (which had no deductible) and the pool’s less expensive and less comprehensive PPO plan (which had a $1,000 deductible). Average per-member-per-month claims were $2,817 for HCTC beneficiaries in the former plan but only $433 in the latter. While additional data may be needed to come to a definitive conclusion, state officials were convinced that adverse selection had taken place, with HCTC beneficiaries who knew their health problems required more expensive care disproportionately tending to enroll in the more comprehensive plan. xi Consumer protections achieved mixed results. Through strenuous efforts by public and private agencies, most workers affected by well-publicized economic displacements made it onto HCTC before 63 day gaps emerged that would have permitted insurers to exclude preexisting conditions. Officials expressed concern that more routine, lower- profile displacements could easily lead to gaps lasting 63 days or longer. HCTC was used by almost no one who experienced gaps that triggered preexisting- condition exclusions, since potential beneficiaries apparently viewed such limited coverage as not worth the cost. Widespread marketing fraud was not a major problem, although there were isolated instances of fraudulent enrollment. HCTC’s centralized enrollment systems prevented the kind of large-scale marketing fraud that was reported for an earlier insurance tax credit program that permitted nongroup insurers to obtain credits by directly recruiting and enrolling potential beneficiaries. However, some nongroup insurers in North Carolina that were not state-qualified plans may have misrepresented themselves to a small number of HCTC beneficiaries and defrauded them of premium payments. In view of these findings, a number of steps are worth serious consideration, both to improve HCTC’s ability to help eligible workers and to enable policymakers to extrapolate from the HCTC experience in designing broader health insurance tax credits. For example, increasing the size of HCTCs (whether for all or some beneficiaries) and new federal strategies to get short-term subsidies to qualified individuals promptly— without asking them to pay monthly premiums in full—will likely be needed to substantially increase enrollment. In addition, further simplification and streamlining of advance-payment procedures may be required to improve take-up and lower administrative costs. Such simplification could give applicants the ability to waive confidentiality, thereby permitting direct communication between multiple public and private agencies that are trying to help particular individuals enroll. Among other benefits, that could permit workers to seek HCTC advance payment by filing one application with one agency, which could then communicate with other entities as necessary. Similarly, careful refinement of consumer protection rules could prove helpful to beneficiaries, government agencies, and health plans alike. It is likewise important to conduct further research that could provide additional information about the problems that have emerged and suggest possible solutions. Only after the HCTC program has been strengthened and tested will it be possible for the HCTC experiment to yield firm conclusions about the inherent capacity of the federal income tax system to subsidize health coverage for the uninsured. xii EARLY IMPLEMENTATION OF THE HEALTH COVERAGE TAX CREDIT IN MARYLAND, MICHIGAN, AND NORTH CAROLINA: A CASE STUDY SUMMARY INTRODUCTION Health policy analysts and political leaders have long proposed using federal income tax credits to help uninsured Americans purchase health coverage. But such mechanisms remained purely theoretical except for a brief and dismal experience in the early 1990s with the so-called “Bentsen Child Health Tax Credits,” a program that benefited few eligible children and generated reports of widespread marketing fraud by insurers that could obtain tax credits through directly recruiting and enrolling potential beneficiaries. This first health insurance tax credit was ended (on a bipartisan basis) after one year of implementation.1 In a very different form, health insurance tax credits were enacted for a second time when President George W. Bush signed the Trade Act of 2002 on August 6, 2002. This legislation created Health Coverage Tax Credits (HCTCs) that pay 65 percent of premiums for certain displaced workers and early retirees who enroll in qualified coverage. Such coverage primarily consists of COBRA plans sponsored by certain former employers and private health plans offered to HCTC beneficiaries by state arrangement. Fully refundable, the credit is available to all who qualify, including those whose income is so low that they owe little or no federal income tax. In a feature IRS officials labeled “revolutionary,” HCTCs can go directly to eligible taxpayers’ health plans each month as premiums come due, in advance of filing annual tax forms. Such “advanceability” was slated to begin on August 1, 2003, slightly less than a year after the law was signed. Despite widespread skepticism about the ability of three cabinet-level departments—Health and Human Services, Labor, and Treasury—to work together effectively and meet that deadline, advance payment of HCTCs began on the specified date, preceded by a one-month “beta test” in several states. In an equally impressive initial achievement, enough states were offering qualified insurance by the end of 2003 that some 75 percent of the country’s HCTC-eligible individuals could actually use HCTCs to enroll in state-arranged coverage.2 Now that these credits, in their advanceable form, have been available for more than 18 months, it is appropriate to describe their early implementation and identify any lessons learned for improving the program and designing broader health coverage 1 expansions. To achieve these goals, this report analyzes early results from three very different states: Maryland, Michigan, and North Carolina. (For a more detailed explanation of HCTC tax credits, see Appendices A and B.) STUDY BACKGROUND AND METHODOLOGY Researchers from the Economic and Social Research Institute (ESRI) began this project by reviewing documents that described early HCTC implementation and by interviewing national policymakers and stakeholders. In April 2004, The Commonwealth Fund and the Nathan Cummings Foundation released the first ESRI report on HCTCs,3 which concluded that federal officials, their private contractors, and state officials had achieved remarkable success at overcoming a range of daunting obstacles and speedily establishing the basic infrastructure of the new program. There were some concerns noted, however, regarding advance-payment enrollment levels, which seemed to indicate that only a small proportion of potentially eligible individuals were taking full advantage of HCTCs. The first report also found that many HCTC health plans treated beneficiaries very differently based on their age, gender, and prior health history; and that the new program’s administrative costs could turn out to be quite high (although some of those costs—related to infrastructure establishment— might eventually be recouped over the lifetimes of both this program and future health insurance tax credit initiatives that build on the HCTC model). The report emphasized, however, that roughly six months after the start of advance payment, it was far too soon to come to any conclusions about the success or failure of HCTCs. In March through May 2004, ESRI researchers visited Maryland, Michigan, and North Carolina to learn more about implementation of HCTCs. These states were chosen because their advance-payment enrollment levels were among the highest in the country (calculated as a percentage of all individuals who were identified as potentially eligible and mailed an HCTC program kit) and because they were diverse in the following ways: the types of state-qualified plans they offered; the relative proportions of beneficiaries who were eligible because of Trade Adjustment Assistance (TAA) and payments from the Pension Benefit Guaranty Corporation (PBGC); their geographic distribution; and the states’ general economic and political characteristics. Through these visits the researchers aimed to learn how—in three very different states—some degree of success had been achieved. The investigators also hoped to identify obstacles to effective implementation that would challenge even the best of public- and private-sector leaders as well as possible strategies for overcoming such obstacles. 2 After examining relevant documents and interviewing representatives of these states’ health plans, governments, labor unions, volunteers, and HCTC beneficiaries, the authors prepared state-specific case studies, which were shared in draft form with state- level informants and are now available, in final form, at www.esresearch.org. This paper summarizes those three case study reports.* OVERVIEW OF IMPLEMENTATION IN MARYLAND, MICHIGAN, AND NORTH CAROLINA Compared with other states, the case study states enrolled into advance payment† a relatively high percentage of individuals who were identified as potentially eligible for HCTC (Figure 1). Figure 1. Advance-Payment Enrollment, Case Study States vs. Other States, September 2004 Percent of potential eligibles who were enrolled in advance payment 15 11.5 9.5 10 7.7 6.1 5.6 5 0 Maryland Michigan North Carolina National States Outside Average the Three Study (including three States study states) Notes: (1) Potential eligibles are individuals who were identified by state workforce agencies (SWAs) or PBGC as potentially qualifying for HCTC and to whom the Internal Revenue Service (IRS) mailed HCTC Program Kits. (2) Because of concerns about confidentiality, the IRS reported enrollment numbers for Maryland and North Carolina that were from 1 to 9 enrollees short of the actual number. The percentages listed here represent the midpoints of these ranges. Source: HCTC Program, Monthly Executive Scorecard, September 2004, October 13, 2004. Calculations by ESRI, November 2004. These states also represented varying profiles in terms of the distribution of PBGC- and TAA-based eligibility among potentially qualifying workers (Figure 2). * This report departs from the state-specific reports by using September 2004 data, rather than May 2004 data, on advance-payment enrollment. † These numbers do not include beneficiaries who claimed HCTCs only on their annual tax forms. Unfortunately, state-specific numbers regarding such claims are not available. 3 Figure 2. Basis for HCTC Eligibility, Case Study States vs. Other States, September 2004 Percent of HCTC eligibles, by basis of eligibility 100 90 PBGC TAA 75 69 61 64 50 39 36 31 25 10 0 Maryland Michigan North Carolina Other States Notes: (1) This figure portrays the basis of eligibility for individuals who were identified by SWAs or PBGC as potentially qualifying for HCTC and to whom IRS mailed HCTC Program Kits. (2) TAA eligibility includes receipt of TAA income support, eligibility for such support but for receipt of unemployment insurance, and receipt of Alternative Trade Adjustment Assistance (ATAA) provided to certain older, displaced workers. Sources: HCTC Program, Monthly Executive Scorecard, September 2004, October 13, 2004. Calculations by ESRI, November 2004. The state-based coverage offered in these three states had several common characteristics, First, higher percentages of HCTC beneficiaries chose such coverage in Maryland, Michigan, and North Carolina than in other states (Figure 3). Figure 3. Advance-Payment Enrollment into Types of Qualified Coverage, Case Study States vs. Other States with State-Qualified Plans, September 2004 Percent of advance payment enrollees using coverage type 100 State-Qualified Plans COBRA Prior Nongroup 80 62 61 57 60 53 46 39 41 36 40 20 1 2 0 2 0 Maryland Michigan North Carolina Other States Notes: (1) Prior Nongroup refers to automatically qualified nongroup coverage that beneficiaries received during at least their final 30 days before job loss or other qualifying event. (2) Because of concerns about confidentiality, the IRS reported that from 1 to 9 individuals were enrolled in prior nongroup coverage in Maryland and North Carolina. The percentages listed here represent the midpoints of each range. (3) Other States includes only states that, in September 2004, offered HCTC beneficiaries state-qualified plans (aside from mini-COBRA plans offered by small firms to former employees, under state law). Sources: HCTC Program, Monthly Executive Scorecard, September 2004, October 13, 2004. Calculations by ESRI, November 2004. 4 Second, although benefit design varied among the three states, each offered HCTC beneficiaries a choice between relatively comprehensive coverage and plans with high deductibles or other significant cost-sharing (Table 1). However, the rules governing premiums varied considerably among the three case study states. Maryland provided a choice between high-risk pool plans, with premiums that varied as a function of age, and a nongroup plan, with “medically underwritten premiums” that took individualized health risk into account. Michigan offered two community-rated plans, neither of which varied premiums based on beneficiaries’ individual characteristics. Finally, North Carolina made nongroup plans available, with medically underwritten premiums. (Table 2). The implications of these variations are discussed below (see finding #7, p. 17). 5 Table 1. Covered Benefits and Cost-Sharing in State-Qualified Plans, Case Study States, June 2004 Maryland Michigan North Carolina Blue Cross/ Blue Blue Plan name MHIP MHIP Blue Shield Community Blue Advantage, Advantage, PPO EPO Personal Comp Blue Value Plan A Plan B High-risk High-risk Nongroup Plan type Community-rated PPO Nongroup PPO pool, PPO pool, HMO indemnity plan $800 (preventive $250; $500; $500; $1,000; Deductible $1,000 None $250 None care exempt) $1,000; $2,500 $2,500; $5,000 Factors affecting premium Age and two tiers Age, gender, area of residence, Age None (other than choice of plan) of health risk status and five tiers of health risk status $15/$30 $25/$50 $20/$30 copay $10 physician 20% co- 50% co- (primary/ (primary/ Network (primary/ visit copay. insurance 25% co-insurance. insurance specialty visit). specialty visit). General specialty visits) 20% other. $10 copay for 20% other care. 30% other care. cost-sharing N/A (no preventive care. Non- 40% co- 40% co- 70% co- 30% co- 30% office visit. non-network network insurance insurance insurance insurance 40% other care. coverage) Guaranteed For enrollees Yes Yes Yes issue? with prior coverage gaps Preexisting No exclusion 6-month exclusion 12-month exclusion conditions? $10/$40 copays 50% copay $2,000 annual Same as plan A, Separate $250 deductible. Prescription $500 annual for network capped at $100 benefit limit. plus separate Must use network pharmacies. drugs benefit limit drugs, 25% for per 34-day $10/$35/$50 $200 $15/$20/$35 copays. non-network supply copays. deductible Maternity Covered without additional Not covered, unless pay Covered without additional premium charge care premium charge extra premium Special limits 60-day inpatient for particular 50% co- 50% coinsurance. 30% co-insurance in network, limit. services insurance for $2,000 annual benefit cap Mental 50% non-network. Coinsurance of 50% co- inpatient care, per person. health Inpatient days capped at 20% for visits 1–5, insurance capped at Lifetime benefit cap of $10,000 60/year. 35% for 6-30, 50% 90 days per person. for 31+ visits. $250 inpatient $500 annual No preventive Preventive care limited to Other copay per limit on care or office network admission preventive care visits 6 Table 2. Cost of State-Based Coverage, One-Person Policies, in Case Study States, June 2004 Beneficiaries’ annualized 35 percent premium share Deductible or 25-year-old woman other major 25- Buying Not buying 55- 55- cost-sharing year-old maternity maternity year-old year-old man coverage coverage man woman MHIP PPO $1,000 $811 $1,415 n/a (policy MHIP EPO None $1,268 $2,557 automatically Maryland BCBS: Cost can includes double, depending on $800 $344 maternity) $667 medical underwriting Community Blue $250 $1,710 n/a (policy $1,710 automatically Michigan includes Blue Value 50% co-insurance $1,292 $1,292 maternity) $250 $483 $1,949 $701 $1,520 $1,575 $500 $458 $1,894 $668 $1,436 $1,487 Plan A North Carolina: Cost can be as $1,000 $424 $1,798 $613 $1,323 $1,369 much as 15 percent lower or up to $2,500 $374 $1,436 $538 $1,151 $1,193 200 percent higher, depending on $500 $391 $1,625 $563 $1,210 $1,252 medical underwriting $1,000 $357 $1,520 $517 $1,109 $1,147 Plan B $2,500 $315 $1,281 $454 $966 $1,000 $5,000 $281 $941 $403 $853 $882 7 FINDINGS The study’s findings fall into three categories: infrastructure, enrollment, and health plans. Infrastructure 1. Public and private entities involved in state-level HCTC implementation were dedicated, creative, and effective. In all three states, gubernatorial leadership was critical in making HCTC implementation a top priority for state agencies, health plans, union officials, and others, many of whom did extraordinary work on this program. For example: • In all three states, state and health plan officials devised creative mechanisms to prevent workers from being required to pay premiums in full during the months when the national program was processing workers’ applications for advance payment. Two states (Maryland and North Carolina) developed innovative “bridge” or “gap-filler” initiatives that used grants from the U.S. Department of Labor (DOL) to pay 65 percent of premiums during those months. Two states (Maryland and Michigan) established mechanisms whereby individuals could delay coverage (and the resulting obligation to pay premiums) until a future date, after the likely start of advance payment. • In North Carolina, clerical staff from Blue Cross and Blue Shield of North Carolina (BCBSNC) volunteered to come in on Sundays to complete the data entry work required for displaced workers to enroll promptly into HCTC coverage. At a factory where major layoffs had occurred, a Community Services Center—staffed by 13 public and private agencies, with government funding supplemented by about $900,000 collected from local organizations—was opened within four days of the plant closing. Local churches “adopted” laid-off workers and helped pay their premiums as HCTC coverage was getting started. A local retiree worked full-time without pay, for months, educating workers and helping them enroll. • In Maryland, one state official was celebrated for his willingness to show up on job sites at any requested hour, including midnight sessions to educate graveyard-shift workers about HCTC benefits and to help them sign up. Another official negotiated details of state-based coverage by phone from the hospital where his wife was in labor with their first child. 8 • In Michigan, one union officer was responsible for helping large numbers of people—union members as well as many others—obtain HCTCs. He worked day and night for several months educating retirees about HCTC benefits, helping them fill out forms and solve emerging problems, and serving as liaison between the applicants and state, national, and health plan officials. • In all three states, government officials worked together effectively across agency boundaries to get HCTC-related systems up and running in remarkably brief periods of time. The level of cooperation and disregard of organizational turf were striking. 2. Agencies involved in national HCTC implementation received generally positive reviews. In all three states, public and private stakeholders expressed gratitude for the very high quality of service they received from the Treasury Department, the Internal Revenue Service (IRS), and the private contractors working with the IRS. Virtually without exception, state-level stakeholders described the staffs of these organizations as responsive, dedicated, and nimble in solving emerging problems. For example, federal officials’ commitment to this program included participation in a number of local community education events to inform workers about available benefits. In another example of their creativity and commitment, federal officials instituted regular monthly reviews of eligibility lists provided by State Workforce Agencies (SWAs). When those lists did not include displaced workers who previously had received advance payment, the officials asked the state agencies to investigate. In many cases, eligible workers had been left off the lists by mistake. This review procedure caught and rectified many such errors, thereby preserving displaced workers’ health coverage. The leadership at DOL, however, was viewed with more ambivalence by state and local stakeholders. On the one hand, officials in several states had to wait many months before receiving DOL responses to their proposed use of NEG grants to pay 65 percent of premiums for workers awaiting IRS rulings on their applications for advance payment. On the other hand, DOL received accolades for ultimately changing its policy in response to unforeseen developments and approving these projects, even though they were not squarely within the agency’s original conception of permitted uses for these NEG funds.4 More broadly, the absence of DOL guidance on a number of issues meant that officials at SWAs were required to devise their own state-specific approaches for dealing 9 with common questions, creating some unnecessary administrative costs. For example, some computer programming had to be done from scratch in each state’s SWA, even though multiple states faced similar problems. A single national model might have greatly reduced the amount of required programming. Feedback on the HCTC program’s national call center varied from state to state, though all informants agreed that the center’s performance improved substantially over the first six months of operation. Nevertheless, the North Carolina interviewees reported that the call center was unable to meet the needs of the state’s HCTC-eligible population. For example, call center staff allegedly used language that beneficiaries found difficult to understand, and they failed to provide the intensive, hands-on assistance needed for many beneficiaries to enroll into advance payment. In the two other states, some beneficiaries found the call center helpful, while others shared the view prevalent in North Carolina. We observed similarly mixed reactions to the outreach materials furnished by the national HCTC program. North Carolina informants uniformly reported that the HCTC Program Kit was far too complicated for beneficiaries to understand. In Michigan and Maryland, by contrast, while some found the kit unintelligible and intimidating, other beneficiaries found it helpful. 3. The process of beneficiary enrollment into advance payment was complex, which increased administrative costs and limited take-up. To benefit from HCTC advance payment, a displaced worker was required to take at least three steps with three different entities: apply at the SWA for TAA assistance,‡ obtain health coverage from a qualified health plan, and apply to the HCTC national office for the start of advance payment. In some cases, displaced workers seeking TAA-related coverage also needed to file a petition with DOL (a fourth agency) seeking a determination that the layoff at issue was trade-related. Moreover, in Maryland and North Carolina, which used DOL funding to pay 65 percent of health insurance premiums pending the start of advance payment, workers seeking such immediate assistance needed to complete an additional application. In Maryland, this extra application had to be filed with an agency that, for most applicants, was distinct from those taking the other HCTC-related applications (in other words, potentially a fifth agency). ‡ This step was not required of PBGC payment recipients, who were mailed HCTC program kits automatically. 10 To further complicate the application process, workers were responsible for passing documentation, in hard-copy form, back and forth between various entities to which they were applying. In all of the case study states, beneficiaries had to provide the HCTC call center with invoices from their qualified health plans. In two of the three states, state- qualified health plans asked workers, before enrollment, to furnish proof that they were on the list of individuals who potentially qualified for HCTCs, either through TAA or PBGC. The two “gap-filler” programs we studied also required hard copies of pertinent documents. When problems have emerged with advance payment, the HCTC program’s general policy has been to require the worker to take the lead in diagnosing and solving them. Despite the excellent coordination among the public and private agencies administering this program at the national level, only rarely have federal, state, and health plan officials been able to communicate directly with one another about the problems encountered by a particular consumer, much less to use automated data exchanges as the principal means of conveying all critical information. This pattern has resulted, in part, from federal officials’ generally laudable commitment to keeping taxpayer information confidential. But it has also required the HCTC program, SWAs, health plans, union staff, and community volunteers to devote tremendous amounts of time and resources to helping individual consumers navigate through this complex system to obtain and retain coverage. For example: • A local SWA office in North Carolina required displaced workers to come to the office to speak by phone with federal officials in front of the SWA staff, who only then could obtain the technical information to help them determine why the workers hadn’t yet received their HCTCs. Federal agency staff would not provide this information directly to SWA officials, even though the beneficiaries had asked them to do so. • In late 2003, when BCBSNC decided to increase premiums beginning in January 2004, the company sought to inform the national HCTC program about these premium changes in advance so future HCTC invoices sent to workers would reflect the correct, revised premiums. Federal officials did not agree to enter the new North Carolina premium amounts into their system, stating that the health plan instead had to educate workers about their new premium costs—and that the workers, in turn, would need to relay their new premium levels to the IRS in a timely fashion. BCBSNC attempted to educate the company’s HCTC members about this procedural requirement, assigning five customer service workers full- 11 time to the task. Despite this effort, many members still did not understand their situation and so did not inform the IRS of the new premium levels. As a result, BCBSNC had to bill these consumers for the full difference between the 2003 and 2004 premiums. Given the resulting confusion and disaffection (together with the higher costs imposed by the new premium levels), 20 percent of HCTC enrollees dropped out of the plan. • The IRS used Social Security numbers (SSNs) to identify HCTC enrollees. Unlike many HCTC plans elsewhere in the country, Maryland’s MHIP (the major state- qualified insurer) used members’ insurance policy numbers rather than SSNs as individual identifiers. To further complicate matters, when the Treasury Department sent money to the insurer’s bank, the standard, national electronic message form did not have room for SSNs and instead used bank account identifiers. As a result, with each month’s payment, MHIP’s third-party administrator went through a lengthy process of reconciling the bank account identifier with the member’s health plan identifier and SSN. The data systems were not coordinated, so most of this work had to be done by hand. • To facilitate HCTC enrollment, SWA staff worked intensively with potential beneficiaries, helping them complete required forms and take other steps to begin and retain advance payment. However, DOL did not generally allow NEG administrative grants to pay for these expenses. Such NEG grants covered data processing, initial infrastructure development, and related activities, but not one- on-one assistance. As a result, SWAs in the case study states had to use other resources to cover some of their HCTC-related administrative costs. • During the first months of the HCTC program, Maryland’s enrollment rate was far higher than that of any other state. One important reason for this success was that PBGC furnished MHIP with a computerized listing of all potentially eligible PBGC recipients in Maryland. With that information loaded into their laptops, MHIP staff could verify eligibility and immediately enroll qualified individuals at outreach events. Because of legal and policy issues concerning beneficiary confidentiality, PBGC later discontinued providing such information, making the enrollment process much more cumbersome. This was one important reason why the state’s enrollment later grew only modestly relative to its early levels. However, steps are being taken to simplify the enrollment process. In several promising developments, the HCTC program has moved forward with confidentiality 12 waivers and electronic provision of information in order to expedite beneficiaries’ enrollment into advance payment. First, in a pilot project with Maryland and Virginia, the HCTC call center has asked callers if their contact information could be shared with state government so that callers can qualify for additional benefits, such as bridge or “gap filler” coverage pending the start of advance payment. Four of five callers (80%) have consented. This innovation led to faster health coverage for eligible workers, increased take-up, and fewer duplicate payments by state bridge programs and HCTC.5 Second, even before the U.S. Government Accountability Office issued a report that addressed this topic,6 HCTC officials agreed that, when premiums are about to change for multiple beneficiaries, health plans can act on behalf of their enrollees in providing the pertinent information to the HCTC program, which will then incorporate that information into its revised billing to beneficiaries. Approximately 30 percent of HCTC premium changes made between March and October 2004 were such bulk premium changes instituted at health plans’ request. If these new procedures had been operative when, in the above example, BCBSNC was preparing to increase premium amounts for January 2004, BCBSNC could simply have given the HCTC program a spreadsheet showing the new premium levels for each beneficiary. This would have greatly lowered the plan’s administrative costs and prevented some disenrollment. Third, in its proposed budget for FY 2006, the Bush Administration has recommended several changes to the HCTC program, one of which “would allow disclosure of certain information necessary to carry out the advance payment program to providers of health insurance or their contractors.”7 As this report goes to press, no additional published information provides more details about this proposal. Nevertheless, all three developments are signposts in the direction of more efficient, consumer-friendly procedures that can reduce administrative costs and increase enrollment and retention. They also serve as another reminder of the responsive and creative administration of this novel program by Treasury, the IRS, and the HCTC program’s private contractors. Enrollment 4. Take-up of advance payment was helped by fortuitous timing, which resulted in favorable publicity and encouraged strong commitments by public and private officials. Each of the three states experienced major economic dislocations as the HCTC program was getting under way. In 2003, Maryland and Michigan both saw Bethlehem Steel, an important employer, terminate pension payments and retiree health coverage, thereby affecting thousands of workers in each state. In Maryland, the state was simultaneously 13 creating its high-risk pool, which further increased media attention both to workers losing health coverage and to HCTCs as a vehicle to subsidize their enrollment into the new high-risk pool. In North Carolina, the Pillowtex manufacturing plant closing represented the largest layoff in that state’s history. Stakeholders in each of the case study states wondered whether the collective response over the long run would be as intense and effective for smaller dislocations that did not receive the same degree of public attention. 5. Most informants agreed that beneficiaries’ inability to afford the 35 percent premium share was the principal reason why so few enrolled in advance payment. According to virtually all informants in North Carolina and Michigan, by far the most important factor limiting take-up of advance payment was that many beneficiaries found their 35 percent premium share to be too expensive. In the words of one health plan official, “affordability is the first, second, and third reason why so few eligible workers enroll.” Maryland was the only state in which the informants did not list affordability as the primary barrier to enrollment. One possible explanation for this difference may be the state’s comparatively higher-income uninsured population. In Maryland, 59 percent of all uninsured individuals had incomes above 200 percent of the Federal Poverty Level (FPL) in 2002,8 compared with 36 percent in the country as a whole.9 A second enrollment barrier mentioned by many informants was the failure of advance payment to begin for several months, consequently obliging beneficiaries to pay their first months’ premiums in full. In the one state we visited (North Carolina) that had fully implemented a bridge or “gap-filler” program to subsidize those initial premium payments,§ stakeholders agreed that the program had a significant positive impact on participation. The third major cause of low take-up, mentioned by numerous informants in all three states, was the complex process required to enroll and remain in advance payment, described above. Of course, an unknown number of potentially eligible individuals did not enroll in HCTC advance payment simply because they received health coverage from other sources, including new employers or spousal employers. The prevalence of such alternative sources of coverage was uncertain, however, and seemed to vary by state. § When we visited Maryland, its “gap-filler” program was just getting started. 14 Informants in North Carolina and Michigan generally believed that most HCTC-eligible workers who did not enroll in advance payment went without insurance coverage altogether. In Maryland, by contrast, interviewees had the impression that many workers who did not receive HCTC advance payment obtained health coverage from other sources. Similarly, the case study states apparently differed in the extent to which beneficiaries had the household resources to pay premiums in full each month during the year and then claim HCTCs on their year-end tax forms. In North Carolina, study informants unanimously reported that only a tiny fraction of HCTC-eligible workers had this capacity. In Michigan and Maryland, it appeared that a larger number of eligible workers, though still a minority, may have been able to file for year-end HCTCs. Some preliminary national data are available that describe year-end HCTC payments for 2003 and that allow some limited, “ballpark” comparison with advance payment enrollment. As of May 2004, 12,594 tax filers had received year-end HCTCs for January through December 2003, based solely on their tax returns, without any advance payment.10 From August through December 2003, 11,917 households enrolled in HCTC advance payment.11 Accordingly, in 2003, the number of households receiving advance payment nearly equaled the number claiming HCTCs only on their annual tax returns. It remains to be seen whether that relationship will hold in 2004 and beyond, however. With advance payment beginning in August but annual refunds available for qualified coverage purchased at any time from January through December, 2003 was a unique year for the HCTC program. We found no evidence supporting two hypotheses that the lead author of this report had previously advanced to explain low take-up.12 First, TAA training requirements did not pose a problem in the three case study states, each of which liberally granted training waivers for HCTC purposes. Training requirements may become more problematic in the future, however, as the TAA statute forbids such waivers for displaced workers after they have received one year of TAA benefits. To obtain HCTCs after the conclusion of that year, displaced workers will need to meet TAA training requirements, without waivers. Regarding the second hypothesis that was not supported by the site visits, SWAs appeared to have effective means at their disposal for reaching HCTC-eligible recipients of unemployment insurance. In the states profiled here, SWAs educated employers and their laid-off workers about the full range of TAA services, including HCTC. Workers 15 could apply for HCTC by filing requests for training, training waivers, or TAA cash assistance. Maryland’s SWA used a particularly effective strategy of supplementing such outreach efforts by querying the state’s unemployment insurance (UI) database to identify UI recipients whose last employer was on the list of Maryland firms that had been certified by DOL as trade-affected. Those recipients were then contacted, informed about HCTC, and given the opportunity to apply. Maryland officials reported that this query was not burdensome to develop or to run on a regular basis, suggesting that other states might profitably consider adopting similar practices. Finally, we likewise found little evidence to support another hypothesis that some have advanced to explain low take-up—namely, that HCTC’s consumer protection requirements deterred health plan participation. Without doubt, those requirements may have reduced the number of participating insurers. It does not follow, however, that the limited number of insurance carriers offering qualified coverage in each state was the primary reason that relatively few potential beneficiaries enrolled. It is true that each state we visited had only one or two carriers participating in HCTC. However, beneficiaries in each state had a number of plans from which to choose, including both comprehensive and high-deductible or other high-cost-sharing options. Blue Cross/Blue Shield, which offered qualified coverage in all three case study states, including the company’s flagship product in North Carolina, had a dominant market share (above 50 percent) in all three states’ nongroup markets.13 Although some federal officials believed that enrollment could have been higher if workers had more flexibility to choose the health plans that qualified for HCTCs, no interviewees at the state level suggested that offering more health plan choices to HCTC beneficiaries would have had a significant effect on take-up. Health Plans 6. Most (but not all) beneficiaries preferred comprehensive benefits, despite the higher resulting premiums. Given a choice between more and less comprehensive coverage, HCTC enrollees into state-based plans typically chose the former: • In Maryland, 60 percent of beneficiaries who enrolled in the Maryland Health Insurance Plan (MHIP) chose the comprehensive HMO option, which had no deductible and cost roughly 50 percent more than MHIP’s PPO, which had a $1,000 deductible. By contrast, among general, unsubsidized enrollees into MHIP, 85 percent chose the PPO. 16 • In Michigan, every advance payment beneficiary enrolled in state-qualified coverage chose Community Blue—which had a $250 deductible, $10 copayments, and 20 percent co-insurance—even though premiums were 32 percent higher than for the Blue Value coverage, which had co-insurance of 50 percent or more and which excluded office visits and preventive care. • In North Carolina, 77 percent of beneficiaries enrolled in the more generous Plan A, and only 23 percent chose Plan B. Between both plans, 52 percent of HCTC beneficiaries selected coverage with $250 or $500 deductibles, rather than plans with deductibles of $1,000, $2,500, or $5,000, even though coverage with lower deductibles cost between 22 and 70 percent more (depending on the beneficiary and the plan involved). Before they learned the results of medical underwriting, even more beneficiaries had enrolled in the most comprehensive options; many switched to less comprehensive choices after receiving individualized (and unexpectedly high) premium quotes. A number of HCTC beneficiaries in North Carolina expressed strong dissatisfaction with the less comprehensive plans that they ultimately selected in order to limit their premium costs. Until deductibles were satisfied, these plans did not cover any medical expenses. As a result, the workers who enrolled in such low-premium coverage reported that they sometimes went without necessary health care. For example, some adults with chronic illness did not refill maintenance medication that had been prescribed to prevent their conditions from worsening. Ultimately, some of these workers perceived their high-deductible coverage as so useless that they disenrolled. 7. Health coverage with pure or modified community rating avoided some serious problems of nongroup coverage that based premium levels on individualized assessment of health risk. One of the most important differences among the case study states involved the possibility of varying premiums based on beneficiaries’ individual characteristics. Michigan’s coverage was purely community rated; for a given plan, each HCTC beneficiary paid the same amount for coverage, regardless of his or her age, gender, area of residence, or prior health history. Likewise, Maryland’s high-risk pool used modified community rating—that is, premiums for each plan varied only by the enrollee’s age. Maryland’s nongroup plan** varied premiums by age and the beneficiary’s placement into one of two risk tiers, based ** At the time of our visit, the high-risk pool served roughly 500 HCTC beneficiaries, and Maryland’s nongroup plan had 7 HCTC enrollees. According to our informants, low enrollment numbers in the nongroup plan resulted mainly from its delayed entry into the market as an HCTC-qualified plan and from its limited coverage of prescription drugs. 17 on an assessment of each individual’s health history and status. North Carolina’s rating rules varied premiums based on age, gender, area of residence, women’s election of coverage for routine maternity care, and placement into one of five risk tiers based on individual medical underwriting. These different rating practices largely reflected each state’s health insurance rules that applied to the general population. Figures 4 and 5 illustrate the extent of premium variation in Maryland and North Carolina for individuals using HCTCs to purchase weighted-average, state-qualified coverage. Michigan’s premiums are not portrayed, since each enrollee was charged the same amount—namely, $1,710 as the beneficiary’s annualized, 35 percent premium share. Figure 4. Annualized 35 Percent Beneficiary Premium Share for Various Individuals Receiving Average State- Based, One-Person Coverage in Maryland, June 2004 Premium share MHIP CareFirst, low risk CareFirst, high risk $6,000 $5,000 $4,000 $3,000 $2,100 $2,100 $2,000 $1,334 $1,334 $1,085 $1,085 $1,000 $688 $688 $667 $667 $344 $344 $0 25-Year-Old Man 25-Year-Old Woman 55-Year-Old Man 55-Year-Old Woman Notes: (1) The average MHIP premium was based on premiums for each age group and plan, weighted based on total HCTC enrollment into both MHIP plans. (2) Only one CareFirst plan was offered to HCTC beneficiaries. The premium amounts for each age and risk combination were provided by CareFirst. Sources: MHIP, CareFirst. Calculations by ESRI, March 2005. 18 Figure 5. Annualized 35 Percent Beneficiary Premium Share for Various Individuals Receiving Average State- Based, One-Person Coverage in North Carolina, June 2004 Premium share Lowest risk Highest risk $6,000 $5,077 $5,000 $3,926 $4,066 $4,000 $3,000 $1,822 $2,000 $1,259 $1,439 $1,112 $1,152 $1,000 $357 $516 $0 25-Year-Old Man 25-Year-Old 25-Year-Old 55-Year-Old Man 55-Year-Old Woman Woman, with Woman, with Maternity Care Maternity Care Included Excluded Notes: (1) This figure shows the premiums that would be charged to various individuals for a single, weighted- average plan, determined based on total HCTC enrollment into all state-qualified options. (2) Lowest-risk and highest-risk premiums were based on risk tier ratios provided by BCBSNC and published quotes for individuals living in the state capital (Raleigh). Source: BCBSNC. Calculations by ESRI, March 2005. These data show that, predictably, premiums did not vary enormously in the two states that used pure or modified community rating. For example, in Maryland: • Gender did not affect premiums. • Individual health risk did not affect premiums charged in the high-risk pool. • Individual health risk could increase premiums by 100 percent for the nongroup plan. • 55-year-olds paid 94 percent more than did 25-year-olds. North Carolina’s premium differences were both more numerous and, in some cases, much larger: • Women age 25 who chose coverage that excluded routine maternity care had premiums that were 45 percent higher than those for 25-year-old men purchasing the same coverage. Among 55-year-olds, female gender had a much smaller impact, raising premiums by only 4 percent. • 25-year-old women who added routine maternity care to their coverage thereby increased premiums by 179 percent. 19 • The impact of age was considerable. Among men, premiums were 211 percent higher for 55-year-olds than for 25-year-olds. Among women, 55-year-olds were charged 123 percent more. • For all ages and genders, placement in the highest rather than the lowest risk tier increased premiums by 253 percent. • Among 25- to 55-year-olds, all these factors combined could raise premiums by 1,322 percent.†† In Michigan and Maryland, neither health plans, state officials, nor other stakeholders expressed dissatisfaction with their states’ rating rules for HCTC coverage. On the other hand, North Carolina’s rating rules created tremendous dissatisfaction, according to virtually every stakeholder interviewed for this report. Many found it unfair that HCTC coverage was least affordable to those who needed it most, with premiums that varied dramatically based on factors outside the individual’s control. North Carolina’s medical underwriting process also had a significant impact on take-up. When they first applied for coverage, HCTC beneficiaries were quoted premiums in the second-best risk tier, as is standard practice for this coverage. At that time, the health plan warned that these quotes were nothing more than estimates and that the actual premium could be higher or lower. After medical underwriting was completed, 70 percent of these applicants saw their premium rates rise above the initial quote, and 69 percent of those in this disappointed group withdrew their applications. If this drop-off had not taken place, more than 3,900 additional North Carolinians could have received advance payment, increasing total national HCTC enrollment by 42 percent.14 These results occurred even though BCBSNC exempted HCTC beneficiaries from the company’s two highest risk tiers, which could have more than doubled premiums above the largest amounts charged to HCTC beneficiaries. If the company had instead applied its standard risk-rating rules, the premium differences, the consequent impact on enrollment, and the level of public disaffection might have been even greater. 8. Some adverse selection and risk segmentation among plan options may have taken place. Because Maryland had been serving HCTC beneficiaries since July 2003 (the earliest among the three case study states), MHIP had sufficient experience to provide ESRI’s †† This represents the difference between premiums for low-risk, 25-year-old men and those charged to high-risk, 25-year-old women choosing maternity care coverage. 20 research team with pertinent claims data. These records showed a sharp difference between HCTC enrollees into MHIP’s more generous HMO (which had no deductible) and those choosing the less expensive but less comprehensive PPO (which had a $1,000 deductible). Between July 2003 and May 2004, average per-member-per-month claims were much higher for HCTC beneficiaries enrolled in the HMO than for those in the PPO ($2,817 vs. $433). Of course, these were raw numbers and not adjusted for enrollees’ age. Nevertheless, state officials were convinced that significant adverse selection and risk segmentation might have taken place in HCTC beneficiaries’ choices among plans, with individuals who foresaw large health care costs disproportionately choosing the more comprehensive coverage.‡‡ 9. Consumer protections in the federal statute had mixed results. Under the Trade Act, if a beneficiary experiences a gap in coverage lasting 63 days or longer before enrolling in a state-qualified plan, the plan can exclude preexisting conditions for up to 12 months or even refuse to issue a policy. The study informants agreed that this posed a considerable challenge, as 63 days or more could easily pass between job loss and the start of HCTC advance payment. For example, because the state agency responsible for unemployment insurance in Michigan had been recently restructured at the time of the site visit and was thinly staffed, it was not unusual for more than 40 days to pass between a worker’s application for TAA and the agency’s full processing of the application. Only after such processing was complete would the Michigan SWA transmit the worker’s information to the national HCTC program, which would then mail the worker information about the HCTC program and encourage the worker to apply for advance payment and enroll in an HCTC health plan. Such processing times compounded other delays, such as that resulting from the statutory prohibition against granting TAA (and HCTC) during the first 60 days after DOL receives a petition seeking a determination of trade-related adverse impact. Notwithstanding these challenges, officials in all three states worked very hard to prevent potential beneficiaries from experiencing gaps in coverage lasting 63 days or more. As a result, most Bethlehem Steel retirees (in Michigan and Maryland) and workers displaced by Pillowtex (in North Carolina) who wished to apply for HCTC coverage could do so before they experienced a 63-day gap. However, in states or communities experiencing smaller economic setbacks that received less attention, the vigorous steps that ‡‡ In contrast, the per-member-per-month cost in MHIP’s general population varied little between HMO and PPO enrollees—respectively $2,326 vs. $2,016. This outcome may have resulted from the explicit limitation of general MHIP eligibility to medically uninsurable individuals, most of whom presumably incur significant ongoing costs, regardless of which plan they choose. HCTC, on the other hand, is open to both the healthy and the chronically ill, so healthy individuals can sort themselves into less expensive plans. 21 retained continuous coverage in high-profile dislocations may not have been as likely, according to study informants. The resulting gaps in coverage would subject workers to preexisting-condition exclusions. Although federal law permits state-qualified plans to deny coverage outright and to impose preexisting-condition exclusions when beneficiaries lack continuous coverage, it does not require them to do so. Accordingly, in the case study states, all state-qualified plans guaranteed the issue of policies to HCTC beneficiaries experiencing prior coverage lapses. Moreover, in Maryland both the high-risk pool and the nongroup plan have been available at various times without any preexisting-condition exclusions for HCTC beneficiaries who had coverage gaps, and neither insurer reported any resulting problems. However, state-qualified plans in Michigan and North Carolina did impose preexisting- condition exclusions for individuals without prior continuous coverage. Such exclusions lasted for six months in Michigan (because of state law) and 12 months in North Carolina. According to study informants, when coverage gaps led to preexisting-condition exclusions, almost no one enrolled into HCTC advance payment. Potential beneficiaries apparently perceived the resulting, limited coverage as not worth the 35 percent premium cost.§§ 10. Widespread enrollment fraud did not materialize, but isolated problems emerged. In all three states, enrollment was centralized. TAA-related individuals had to go to their SWA, and all HCTC beneficiaries seeking advance payment needed to enroll in a qualified plan and contact the national HCTC office. These centralized points of entry made it difficult for rogue insurers to engage in marketing abuses, such as those widely reported in connection with the Bentsen Child Health Tax Credits in the late 1990s.15 Under that earlier program, insurers could market themselves directly to potential beneficiaries and obtain their tax credits by enrolling them into coverage. By avoiding such open-ended enrollment options under the direct control of insurers, the HCTC program prevented the recurrence of similar large-scale abuses in the three study states. However, some nongroup insurers in North Carolina that were not state-qualified plans might have misrepresented themselves to a small number of HCTC beneficiaries and §§ Officials in Virginia came to the similar conclusion that “timeliness is everything.” In that state, direct mail obtained only a 5 percent response rate when it was sent to individuals receiving benefits who, for the most part, had prior coverage lapses. When face-to-face rapid response took place before the 62-day window closed, more than 66 percent of potential beneficiaries responded. Health Coverage Tax Credit Program. HCTC-NEG Bridge Pilot. July 2004. 22 defrauded them of premium payments, according to observers. When workers laid off by Pillowtex were congregating at the mill to receive training and assistance regarding available benefits, some nongroup insurers that did not offer state-qualified coverage placed tents nearby. Agents of these plans apparently urged some of the laid-off workers in attendance to write their first month’s premium checks on the spot, falsely promising that HCTCs would reimburse the workers later. This alleged abuse seemed to represent an isolated occurrence, but it provided an important reminder of the problems that could become widespread in a more weakly structured environment. AREAS FOR POTENTIAL IMPROVEMENT This report seeks to provide information that will allow policymakers to achieve two different objectives: namely, to enhance the help that HCTCs provide to their current target population; and to use the HCTC experience to guide the future design of measures to cover other, potentially much larger groups of uninsured Americans. To reach these goals, the following policy approaches may be worth considering: 1. Increase the percentage of premium covered by HCTCs. This can be done for either all beneficiaries or those with low incomes. Unless this obstacle to enrollment is addressed, take-up is likely to remain low, according to most of the study informants. 2. Simplify and streamline advance-payment procedures for beneficiary enrollment. If the IRS developed procedures that permitted individuals to apply for HCTC advance payment by submitting a single form to a single entity, administrative costs could fall and take-up rates could rise. Such procedures could include the applicant’s waiver of privacy rights under the tax code, the PBGC program, the Health Insurance Portability and Accountability Act (HIPAA), and other pertinent laws. With such comprehensive privacy waivers, all entities involved in HCTC—health plans, their third-party administrators, the national HCTC program, SWAs, and PBGC—could share information (via automated data exchange, whenever possible) to ensure the timely receipt of coverage by individuals who legitimately qualify for credits. 3. Provide immediate assistance. Many beneficiaries lack the excess income in household budgets to pay even one or two months’ premiums in full. To address the need for immediate subsidies once workers have lost prior coverage, bridge or “gap-filler” programs like those in North Carolina and Maryland could be instituted in additional states as a stopgap measure. More 23 comprehensively, Congress could authorize the IRS to provide a short-term, advanceable credit payable immediately upon certification by a designated agency that individuals are enrolled in qualified health plans and that they are on eligibility lists transmitted to the IRS by an SWA or PBGC. Such designated agencies could include qualified health plans and SWAs. 4. Fund case management services for enrollment and retention. Until the advance-payment system is significantly streamlined, many HCTC beneficiaries will need a good deal of assistance enrolling in and retaining advance payment. A variety of public and private entities, including SWAs, health plans, unions, the HCTC call center, and consumer advocates, have the potential capacity to provide such assistance. However, financial support is required for this capacity to be realized on an ongoing basis. If present NEG grant amounts and IRS administrative appropriations are insufficient, more resources should perhaps be provided, through either these or other funding streams. 5. Address concerns about consumer protections. Congress could revise the HCTC consumer protection rules to disregard, for purposes of determining whether an individual has had continuous coverage, two periods of time: the interval between job loss or other qualifying event and an individual’s receipt of notice of potential HCTC eligibility; and the interval between an individual’s application for advance payment and the start of advance payment. This approach would seek to accomplish two goals at once: first, preventing administrative delays from causing essentially involuntary gaps in coverage that terminate consumer protections; and second, withholding consumer protections from any beneficiaries who deliberately wait until they get sick before using their HCTCs to enroll in coverage. 6. Address concerns about nongroup coverage with medically underwritten premiums. It is premature, based on these case studies alone, to come to broad conclusions about tax credits used for nongroup coverage. Nevertheless, North Carolina’s problematic experience suggests an urgent need to conduct further research that explores the consequences of using HCTCs to subsidize nongroup coverage with medically underwritten premiums that vary substantially based on individual risk. 7. Expand the kinds of research done by the IRS. To gain a broader understanding of enrollment trends for advance payment, the IRS could analyze aggregate information from filed income tax forms for 2003 to compare the characteristics of five types of individuals: those who were mailed HCTC Program Kits 24 but did not apply for advance payment; those who attempted to enroll in HCTC advance payment but did not complete that process; those who successfully enrolled in HCTC advance payment; those who filed for HCTCs on their annual tax form exclusively; and those who obtained HCTCs through both advance payment and refund claims on annual tax forms. If used in aggregate form to respect the confidentiality of individual taxpayer information, this analysis could include income,*** which could potentially provide important information about the role of affordability in limiting take-up. In addition to analyzing tax return information, the IRS could conduct a new round of HCTC surveys. These might investigate, among other things, the reasons why individuals who were mailed HCTC Program Kits did not enroll, including the proportion who received coverage from other sources and therefore may not have needed or qualified for HCTCs. The surveys could also evaluate access to health care and degree of satisfaction with such care among potential beneficiaries who enrolled in various forms of HCTC-qualified coverage; those who were uninsured; and those who received coverage from other sources (without using HCTCs). One final category of needed research would require congressional authorization. The IRS could be empowered to conduct demonstration projects that involve changes to the HCTC program. Such authorization could be quite open-ended, modeled after Section 1115 of the Social Security Act. This would permit the IRS to investigate and test a broad range of hypotheses and alternative strategies that could result in substantial improvements in program design, as well as in greatly enhanced knowledge about the capacity of tax credits to assist individuals losing health coverage. IMPLICATIONS FOR STRUCTURING BROADER COVERAGE EXPANSIONS Based on the HCTC experience to date, we do not yet know whether income tax credits for the purchase of health coverage can achieve the following: • use of efficient, consumer-friendly methods for enrolling into advance payment without multiple applications, extensive paperwork requirements, and significant beneficiary confusion; • rapid commencement of advance payments by the federal income tax system, without requiring beneficiaries to pay full monthly premiums while the IRS is determining their eligibility for advance payment; and *** For individuals who neither claimed HCTCs nor filed annual income-tax returns, such non-filing could serve as a proxy indicator that household income may have been low. 25 • limitation of administrative costs to reasonable levels. Even if credits can in principle accomplish these goals, important policy design questions need to be answered—on the basis of real-world experience—before policymakers can be confident of success in the expansion of tax credits to serve additional populations. Here are some of the key questions that could benefit from further research into the HCTC program: • What level of subsidy and what enrollment mechanisms lead to significant take-up by eligible individuals who would otherwise be uninsured? • What trade-offs result from using tax credits to subsidize enrollment in nongroup plans with medically underwritten premiums that vary greatly based on individual health risks? • What level of adverse selection and risk segmentation is likely to result from diverse health plan offerings open to tax credit beneficiaries? How can the extent or harmful consequences of such segmentation be reduced? CONCLUSION While it is still too soon to reach a verdict on the nation’s second experiment with using the federal income tax system to subsidize health coverage for the uninsured, the good news is that thousands of vulnerable individuals have received essential assistance. Moreover, officials at all levels, public and private, have done extraordinary work to establish the new system’s infrastructure, to prevent significant marketing fraud, and to make the existing policy as useful as possible to displaced workers and early retirees who need and qualify for help. However, significant problems have emerged. If policymakers wish to establish the viability of tax credits for expanding health coverage for millions of uninsured Americans, it is important to conduct further research that will analyze emerging problems with HCTCs and identify potential solutions. At a fundamental level, only if HCTCs are adjusted to be as effective as possible can policymakers use the HCTC experience to gauge the inherent capacity of tax credits to cover the uninsured. 26 APPENDIX A. NATIONAL HCTC RULES Comprehensive explanations of the Health Coverage Tax Credits (HCTC) program are available elsewhere.16 For the purposes of this report, however, the following brief summary may be helpful: Eligibility. Several groups qualify for HCTCs: • Displaced workers whose layoffs have been certified by the U.S. Department of Labor (DOL) as trade-related and who therefore receive Trade Adjustment Assistance (TAA) cash payments or would qualify for such payments but for their receipt of unemployment insurance (UI); • Certain adults aged 55 through 64 who are paid by the Pension Benefit Guaranty Corporation (PBGC), which assists retirees from companies that have suffered severe financial reversals and so no longer pay promised defined benefit pensions; • Adults aged 50 through 64 who receive Alternative Trade Adjustment Assistance (ATAA) payments because they suffered trade-related job loss and then shifted to a new line of work for lower pay; and • Dependents of individuals in the three previous categories. Individuals must also meet other criteria for eligibility, including absence of health coverage through Medicare, Medicaid, or employer-based plans (either as a worker or dependent) for which the firm pays 50 percent or more of premiums. Health coverage. HCTCs pay 65 percent of premiums for qualified health plans, which fall into two categories: • State-qualified coverage, which is established by state action (through arrangements with an insurer or certain other methods) and which must meet the consumer protection requirements described below; and • Automatically qualified plans, which are available for HCTC use throughout the country (without any required action by state government) and which include: (a) COBRA plans offered by former employers; and (b) nongroup coverage in which the HCTC beneficiary was enrolled during at least the final 30 days before job loss or other qualifying event. 27 Consumer protection requirements. For individuals with at least three months of continuous coverage immediately before enrolling in an HCTC plan, without any gap in coverage of 63 days or more, state-qualified insurers must guarantee issue and may not exclude coverage of preexisting conditions. Modes of obtaining HCTC. An eligible individual may either claim the HCTC on annual tax returns (for reimbursement of insurance premiums paid during the year) or have the HCTC paid in advance to the insurer, each month, as premiums are due. HCTC-related grants to states. These grants, made by DOL and the Centers for Medicare and Medicaid Services (CMS), fall into two categories: CMS grants to support the establishment and operation of high-risk insurance pools; and National Emergency Grants (NEGs) from DOL, which include funding for infrastructure development and certain transitional state costs associated with HCTCs. 28 APPENDIX B. ENROLLMENT INTO ADVANCE PAYMENT While each state tailors the process in accord with its own infrastructure, and the national program makes adjustments on an ongoing basis with the goal of improving outcomes for beneficiaries, taxpayers, and health plans, enrollment into HCTC payment generally includes the following steps: 1) Each State Workforce Agency (for TAA and ATAA beneficiaries) and the Pension Benefit Guaranty Corporation (for its beneficiaries) send the HCTC program lists of individuals who may qualify for HCTC. Transmitted electronically, such lists are provided daily by SWAs and monthly by PBGC. 2) The HCTC program mails HCTC Program Kits to each individual listed by the PBGC or an SWA as potentially eligible. These kits contain detailed explanations of eligibility, qualified coverage, application procedures for HCTCs, and related topics. They are available in English and Spanish. 3) The individual enrolls in qualified health coverage. Unless the potential beneficiary lives in a state offering so-called “bridge” or “gap-filler” assistance, he or she must pay each month’s premium in full, pending completion of the HCTC registration process and the start of advance payment. If the individual turns out to be eligible for HCTCs, the Internal Revenue Service (IRS) reimburses such full-premium payments at the end of the year, after the individual files annual income tax forms. 4) The health plan sends an invoice to the individual showing the full premium amount. (The plan can continue sending these full, premium invoices even after advance payments begin and the consumer makes 35 percent premium payments to the HCTC program, rather than pay full premiums to the plan.) 5) The individual contacts the HCTC call center to enroll in advance payment, while mailing the health plan’s invoice to the HCTC program. The HCTC program then uses that invoice to confirm enrollment in qualified coverage and to determine the proper dollar amount of the credit and the regular, monthly due date for payment to the plan. 6) The HCTC program determines whether the individual is eligible and, if so, registers him or her for advance payment. 7) For each month of advance payment, the following process applies: a) Precisely 27 days before the plan needs to receive its full premium payment, the HCTC program bills the individual for his or her 35 percent premium 29 share. The consumer’s payment is due at the HCTC program 21 days after the bill is mailed. b) If the HCTC program receives the full 35 percent payment by that date, the IRS provides a 65 percent advance credit. The HCTC program then combines that credit with the beneficiary’s payment, sending the full premium payment electronically to the health plan. c) If the beneficiary pays less than the full 35 percent amount, the HCTC program combines that payment with a proportional matching credit from the IRS, forwards the combined payment to the health plan, and reminds the beneficiary of the additional amount that must be paid to the plan to retain coverage. The plan is then responsible for collecting such additional amounts. d) If no payment is received from the beneficiary by the due date, the HCTC program sends the consumer a notice stating that HCTC neither received a payment from the consumer nor made a payment to the health plan on the consumer’s behalf—and that to maintain coverage the consumer must pay the full premium amount to the plan. As long as it continues to receive eligibility records from PBGC or an SWA, HCTC continues sending advance-payment invoices to the consumer requesting 35 percent of the next month’s full premium amount. It is up to the individual and the plan to work out issues regarding past amounts due and whether coverage continues. Eligible individuals who are enrolled in qualified plans and who make premium payments that are not covered by advance payment may use their annual tax returns to claim HCTC reimbursement for such payments. 30 NOTES 1 House Ways and Means Committee, Subcommittee on Oversight. Report on Marketing Abuse and Administrative Problems Involving the Health Insurance Component of the Earned Income Tax Credit. WMCP: 103-14, 103rd Cong., 1st Sess., June 1, 1993. 2 HCTC Program. Monthly Executive Scorecard, December 2003. January 20, 2004. 3 Stan Dorn and Todd Kutyla. Health Coverage Tax Credits Under the Trade Act of 2002. Economic and Social Research Institute, The Commonwealth Fund. April 2004. http://www.cmwf.org/usr_doc/dorn_725_trade_act.pdf. 4 Assistant Secretary Emily Stover Derocco. Training And Employment Guidance Letter No. 20-02, Change 1. May 13, 2004. Employment and Training Administration Advisory System, U.S. Department of Labor. http://www.ows.doleta.gov/dmstree/tegl/tegl2k2/tegl_20-02c1.htm. 5 HCTC Program. HCTC-NEG Bridge Pilot. July 2004. (Results as of April 30, 2004). Calculations by ESRI, February 2005. 6 Government Accountability Office. Health Coverage Tax Credit: Simplified and More Timely Enrollment Process Could Increase Participation. GAO-04-1029. September 2004. http://www.gao.gov/new.items/d041029.pdf. 7 U.S. Department of the Treasury. General Explanations of the Administration’s Fiscal Year 2006 Revenue Proposals. February 2005. http://www.treas.gov/offices/tax-policy/library/bluebk05.pdf. 8 Maryland Health Care Commission. Health Insurance Coverage in Maryland Through 2002. November 20, 2003. http://www.mhcc.state.md.us/health_care_expenditures/insurance_coverage/mhcc_insurance_report_1103.pdf. 9 Catherine Hoffman and Marie Wang. Health Insurance Coverage in America: 2002 Data Update. December 2003. Kaiser Commission on Medicaid and the Uninsured and The Urban Institute. http://www.kff.org/uninsured/4154.cfm. 10 HCTC Program. Monthly Executive Scorecard, September 2004. October 13, 2004. 11 The Lewin Group. Advance Premium Payments: A Snap Shot of Early Experience. Data from December, 2003. January 2004. 12 Dorn and Kutyla, op cit. 13 In 2001, the latest year for which published, national data are available, Blue Cross/Blue Shield covered 52 percent, 66 percent, and 59 percent of all nongroup market enrollees in Maryland, Michigan, and North Carolina, respectively. Deborah Chollet, Fabrice Smieliauskas, Madeleine Konig. Mapping State Health Insurance Markets, 2001: Structure and Change. Mathematica, Inc., for State Coverage Initiatives Program, Robert Wood Johnson Foundation. September 2003. http://www.statecoverage.net/pdf/mapping2001.pdf. 14 HCTC Program. Monthly Executive Scorecard, January 2004. February 25, 2004. Calculations by ESRI, November 2004. 15 House Ways and Means Committee, Subcommittee on Oversight, op cit. 16 Stan Dorn. The Trade Act of 2002: Coverage Options for States. Economic and Social Research Institute, for AcademyHealth’s State Coverage Initiatives Program, March 2003, http://www.statecoverage.net/pdf/issuebrief303trade.pdf. Official, detailed explanations of Trade Act health coverage are available online, including at http://www.irs.gov/pub/irs-utl/governers_letter_hctc_guidance_ltr_ammended_080803_v2.pdf and http://www.irs.gov/individuals/article/0,,id=109960,00.html. 31