The creation of health insurance exchanges. How Choices In Exchange Design For States Could Affect Insurance Premiums And Levels Of Coverage

Transcription

1 velop and operate exchanges on their own and for those choosing to develop and operate exchanges jointly with the federal government. 1,2 The act s flexibility allows each state to tailor its exchanges to the characteristics and preferences of its population. For example, states can decide how their exchanges will be structured and governed; the extent to which the exchanges will operate as active purchasers of insurance; and which insurance rating rules will apply to private insurance markets operating both inside and outside the exchanges, within boundaries set by the federal government. States will determine whether to standardize plan offerings within the federal acdoi: /hlthaff HEALTH AFFAIRS 31, NO. 2 (2012): Project HOPE The People-to-People Health Foundation, Inc. By Fredric Blavin, Linda J. Blumberg, Matthew Buettgens, John Holahan, and Stacey McMorrow How Choices In Exchange Design For States Could Affect Insurance Premiums And Levels Of Coverage Fredric Blavin (fblavin@ urban.org) is a research associate at the Health Policy Institute, in Washington, D.C. Linda J. Blumberg is an economist and senior fellow at the Urban Institute Health Policy Center. Matthew Buettgens is a senior research associate at the Urban Institute Health Policy Center. John Holahan is director of the Urban Institute Health Policy Center. Stacey McMorrow is a research associate at the Urban Institute Health Policy Center. ABSTRACT The Affordable Care Act gives states the option to create health insurance exchanges from which individuals and small employers can purchase health insurance. States have considerable flexibility in how they design and implement these exchanges. We analyze several key design options being considered, using the Urban Institute s Health Insurance Policy Simulation Model: creating separate versus merged small-group and nongroup markets, eliminating age rating in these markets, removing the small-employer credit, and setting the maximum number of employees for firms in the small-group market at 50 versus 100 workers. Among our findings are that merging the small-group and nongroup markets would result in 1.7 million more people nationwide participating in the exchanges and, because of greater affordability of nongroup coverage, approximately 1.0 million more people being insured than if the risk pools were not merged. The various options generate relatively small differences in overall coverage and cost, although some, such as reducing age rating bands, would result in higher costs for some people while lowering costs for others. These cost effects would be most apparent among people who purchase coverage without federal subsidies. On the whole, we conclude that states can make these design choices based on local support and preferences without dramatic repercussions for overall coverage and cost outcomes. The creation of health insurance exchanges is a centerpiece of the health reforms being implemented under the Affordable Care Act of 2010, along with expansion of the Medicaid program, reform of insurance-market regulations, and provision of subsidies to make private coverage more affordable. Exchanges will serve as organized marketplaces for the purchase of insurance in the individual and smallemployer markets. The planning and development of these exchanges is currently the focus of substantial state and federal resources. The Affordable Care Act provides a number of design choices both for states choosing to de- 290 Health Affairs February :2

2 tuarial value tiers and the degree to which insurance plans participating in the exchanges will be monitored to ensure compliance with national and state rules. States must provide certain types of information to consumers about the insurance plans offered in the exchanges for comparison purposes, such as information about enrollment and disenrollment, claims denials, and rating practices. They do, however, have flexibility in how that information is conveyed to the public, and how consumer friendly it is. The choices states make in designing their exchanges will have implications for the size of the exchanges in each state, both the individual-market exchanges those set up for the nongroup market and the Small Business Health Options Program (SHOP) exchanges those set up to provide insurance plans for smallbusiness owners and their employees. The size of the exchanges, in turn, will affect each state s leverage in negotiations with providers. Ultimately, these choices will affect the characteristics of those enrolled in the exchanges, the premiums that different types of consumers will pay, the distribution of health insurance coverage, and the cost to the government. Policy makers need to understand the trade-offs associated with the many options in order to make informed decisions about an array of exchangerelated issues. In this article we use the Urban Institute s Health Insurance Policy Simulation Model to explore the implications of a selected number of state exchange design choices. These choices are summarized below. Design Choices For States Separate Versus Merged Markets States have the option of merging their small-group and individual insurance markets inside and outside the exchanges. Merged markets would be treated as one combined risk pool for premium rating purposes, and risk adjustment would be performed across all plans in the combined markets inside and outside the exchange. Age Rating Bands Of 3:1 Or Less The Affordable Care Act requires that premiums in the small-group and individual markets vary by no more than a 3:1 ratio based on age. This means that an older adult purchasing coverage in these markets cannot be charged a premium that is more than three times the premium charged to a younger adult for identical coverage. States can, however, set the number of age bands and intervals, and they are permitted to limit agedependent premium variations even further than the law requires for example, by requiring plans to charge all adults the same premium for the same coverage, regardless of age. Small-Employer Tax Credit The Affordable Care Act establishes a small-employer tax credit that partially offsets the premium costs associated with coverage provided by small, low-wage employers. This credit has been available since 2010 and will continue through Starting January 1, 2014, the credit rate will increase. The credit will be available to eligible small employers for a total of two consecutive years from that date forward. The true effect of the credit after 2014 will fall somewhere between the credit and no-credit scenarios, because there will be a mix in the market of those who have exhausted their time-limited credit eligibility and those who have not. Although the tax credit does not represent a design choice for states, we simulate the difference between the credit and no-credit scenarios as a sensitivity analysis. Definition Of Small Group Market Prior to 2016, states may define small group as employers with fifty or fewer workers. Beginning in 2016, the definition must be employers with 100 or fewer workers. The definition determines which employers are eligible to purchase coverage for their workers through the SHOP exchange. In addition, the Affordable Care Act contains an array of new insurance market reforms that apply solely to the small-group and individual markets. Consequently, the definition of small group also determines which insurance markets will be affected by these reforms between 2014 and Reforms to small-group and individual markets include prohibitions on health status rating, age rating limits of no more than a 3:1 ratio as described above, tobacco use rating limits of no more than a 1.5:1 ratio, mandatory provision of essential health benefits, a requirement that plans be provided within specified actuarial value tiers, and risk adjustment across plans. Study Data And Methods We used the Health Insurance Policy Simulation Model to estimate the effects of health reform, simulating the decisions of businesses and individuals in response to policy changes such as the Medicaid expansion, new health insurance options, subsidies for the purchase of health insurance, and insurance market reforms. The model provides estimates of changes in the following outcomes: government and private spending, premiums, rates of employer offers of coverage, and health insurance coverage resulting from specific reforms. February :2 Health Affairs 291

3 The core of the model is the Current Population Survey Annual Social and Economic Supplement, matched to several other national data sets, including the Household Component of the Medical Expenditure Panel Survey. We used detailed information on state eligibility rules to simulate Medicaid eligibility and enrollment, 3 and we made further adjustments to account for the documented undercount of Medicaid enrollment in the Current Population Survey and other household surveys. The model also simulates the documentation status of immigrants, allowing for more accurate identification of those eligible for public programs and exchange enrollment. 4 The model generates a set of synthetic firms that tracks the national distribution of employers of different types. It also populates these simulated employers with workers employed in the same types of firms. In this way, we can base the decisions of employers to offer coverage to workers upon the characteristics of the workers in each firm and their dependents. Premiums in employer and individual insurance risk pools are computed as a function of the medical expenses of those enrolled, administrative costs, and any relevant regulatory rules. The model adjusts premiums to reflect available government subsidies for particular populations and coverage options and can respond to a variety of regulatory changes and types of enrollment behavior. Individuals and families decisions in the model are based in a flexible economic expected utility framework. The value of each coverage option available to each individual or family is a function of the following: the direct premiums they pay, the value of the health care they use, their expected out-of-pocket health expenses, the variance of out-of-pocket expenses (a measure of risk), any premium or cost-sharing subsidies or tax incentives for which they are eligible, the individual s or family s expected outof-pocket expenses relative to income, and any incentives to purchase created by the individual coverage requirement and the new employer assessments. Individuals choose the insurance option, including the option of remaining uninsured, that carries the highest valuation for them. Overall price elasticities of employer and household behavior are calibrated to the strongest empirical economics literature. 5 7 Simulations We simulated the main coverage provisions of the Affordable Care Act as if they were fully implemented in 2011, for ease of comparison across the options. It is important to note, however, that the model s results were computed in an iterative process, with the final results reflecting a long-run equilibrium outcome. 8 We compared each design option to a standard implementation of the Affordable Care Act, as described below. SIMULATION 1(STANDARD CASE): This simulation assumes the following: First, the smallgroup and nongroup markets are kept separate for rating purposes; second, small group is defined as 100 or fewer employees; third, there is 3:1 age rating; and fourth, the small-employer tax credit is available to all small employers offering coverage through the SHOP exchange that meet the employer size and average wage requirements. SIMULATION 2: This simulation differs from the standard case in the first assumption: The small-group and nongroup markets in and out of exchanges are merged for premium-rating purposes. All other assumptions are the same. SIMULATION 3: This simulation differs from the standard case in the third assumption: No age rating is permitted all adults face the same premium in the small-group and individual markets for identical coverage. Tobacco-use rating continues, however. All other assumptions are the same as in the standard case. SIMULATION 4: This simulation differs from the standard case only in the fourth assumption, in that the small-employer tax credit is eliminated. SIMULATION 5: This simulation differs from the standard case only in the second assumption: The maximum number of employees for firms in the small-group market is set at 50 workers. Limitations A number of limitations to our approach are worth keeping in mind. First, modeling full implementation of the Affordable Care Act in a single year, as we do here, makes comparison of the design options straightforward. However, this strategy and the subsequent results do not reflect the likely phasing in of program participation or the effect of cost containment measures, such as bundled payment approaches or accountable care organizations, which will increase over time. Second, behavior within our model is calibrated to the best empirical economic literature on employer and household responses to price changes and the availability of new coverage options. However, certain behavioral decisions are less predictable than others. Specifically, the split between exchange and non-exchange enrollment in small-group coverage carries particular uncertainty. Employers decisions are modeled here as if eligible employers are essentially neutral between SHOP and non-shop coverage at the same price, but the actual decision by small employers will depend upon a number of un- 292 Health Affairs February :2

4 knowns. One such unknown is how small-group plans will differentiate their offerings inside and outside the exchanges. States can require that the offerings be uniform, but this is not required by the Affordable Care Act. Other unknowns include whether states will make all regulatory rules in and out of the exchange uniform in this market, the effectiveness of the risk-adjustment methodology, and the role of brokers. These and related issues are discussed further in the article by Jon Kingsdale in this issue of Health Affairs. 9 In addition, although the model takes into account the administrative costs associated with different insurance options, it does not account for employer search costs for insurance plans. To the extent that the SHOP exchange can substantially decrease such costs, employers with fifty or more workers may find the SHOP exchange much more attractive than is modeled here. At this time, the model does not incorporate changes in employers contributions to workers coverage or an employee-choice option in the SHOP exchange. Also, the simulations presented assume a fully effective risk-adjustment system, but the actual system is likely to fall short of that ideal. Finally, because the regulations associated with the Affordable Care Act are being released on a rolling basis, some uncertainties about the final rules remain. If the final rules differ from expectations, then the results could be affected. One example is the final treatment of affordability computations, subsidy eligibility, and penalty exemptions for family members of workers with affordable employer-based insurance offers. Here we have simulated results using the interpretation of the Joint Committee on Taxation, which says that the affordability exemption and subsidy eligibility are based on single coverage. 10 Study Results Merging The Small-Group And Individual Markets Each policy we examine is compared to the standard case with separate risk pools for the small-group and individual markets and exchanges. Exhibit 1 shows the impact of merging these two markets, creating one risk pool. Merging the small-group and individual risk pools leads to a substantial decline in individual premiums, by more than 10 percent ($600) in the exchange and by somewhat less outside of the exchange. This is due to the combining of the lower-cost small-group pool with the individual pool. Premium differences inside and outside the exchange still exist because of age and tobacco-use rating. For example, individuals in the nongroup market outside the exchange are younger and are less likely to use tobacco than individuals inside the exchange. Premiums in the small-group market, on the other hand, are largely unchanged compared to the standard case. Although we might expect an increase in small-group premiums following the merger, the simulation takes into account complex dynamics that can result in higher or lower premiums. In this case, an increased attractiveness of the nongroup market relative to the group market may alter the preferences of certain employers and their workers, which can ultimately lead some people to shift from small-group to individual coverage. The model accounts for such behavior by individuals and employers, and the resulting small-group premiums reflect these decisions. Compared to the standard-case simulation, merging the markets increases the number of people purchasing coverage through the nongroup exchange by about 1.7 million. The largest part of the increase in this exchange comes from a movement of about 910,000 people from employer-sponsored insurance outside the exchange, as a result of a slightly lower offer rate. Because employer-sponsored insurance premiums for small firms are higher when the risk pools are initially merged, more employees will find individual coverage attractive than when the markets are separated. This applies to employees regardless of their eligibility for subsidies in the nongroup market. Moreover, as nongroup premiums decline substantially, nongroup coverage becomes even more attractive for those not receiving subsidies. Overall, a merged-markets scenario results in about one million fewer uninsured people than in the case of separate risk pools. A sizable portion of the increase in nongroup exchange participation occurs because nongroup coverage becomes more affordable and more of the uninsured decide to obtain coverage. Medicaid rolls also increase, somewhat counterbalancing the higher numbers of uninsured people caused by the drop in employer coverage. That is, some of those affected by the reduction in the number of employers offering employer-sponsored insurance are eligible for coverage through Medicaid and the Children s Health Insurance Program and so opt to enroll in those programs when they are not offered an employer option. Premium subsidies decline when markets are merged. Premium subsidies are lower because the blending of the risk pools causes premiums in the individual exchanges to fall. Cost-sharing subsidies increase modestly as a result of some increased enrollment by the low-income workers February :2 Health Affairs 293

5 Exhibit 1 Effects Of Merging Small-Group And Individual Markets On Insurance Coverage And Costs, Compared With Standard-Case Assumptions, 2011 Standard case: separate markets Option: merged markets Number Percent Number Percent Difference Total population (millions) a a a Insured (millions) Exchange Employer group Individual Subtotal Non-exchange Employer group Individual Medicaid Other Uninsured Premiums per covered life: small firms Exchange $4,800 a $4,800 a $0 Non-exchange 4,900 a 4,900 a 0 Total 4,900 a 4,900 a 0 Premiums per covered life: nongroup Exchange $5,700 a $5,100 a $600 Non-exchange 5,100 a 4,700 a 400 Total 5,700 a 5,100 a 600 Subsidies (billions of dollars) Premium subsidies $33 a $29 a $4 Cost-sharing subsidies $5 a $5 a $0 Employer subsidies $4 a $4 a $0 Employer spending (billions) Net employer spending $541 a $537 a $4 Employer-group premiums 541 a 537 a 4 Employer assessments 3 a 3 a 0 Employer subsidies 4 aa 4 a 0 Individual spending (billions) Total individual spending $371 a $372 a $0 Spending by those below 200% of poverty 54 a 53 a 1 Spending by those at % of poverty 112 a 113 a 1 Spending by those at 400% of poverty or more 206 a 206 a 0 SOURCE Urban Institute analysis, Health Insurance Policy Simulation Model, NOTES We simulated the provisions of the Affordable Care Act fully implemented in Premiums are rounded to the nearest hundred dollars. a Not applicable. who have lost their employer-based coverage. Exhibit 1 also shows that despite no change in premiums per covered life in the small-group market, overall spending by employers on premiums falls slightly when markets are merged. This is attributable to the decline in the number of people purchasing employer-sponsored coverage. There is little difference in spending by individuals and families by income group between the two scenarios. Elimination Of Age Rating The goal of eliminating age rating is to spread health risk more evenly across young and old, lowering premiums for older age groups while increasing them for younger age groups.we find that eliminating age rating in the small-group and individual markets increases individual premiums for single policyholders ages by $1,400, from $3,600 to $5,000, and for single policyholders ages by $800, from $4,200 to $5,000. However, premiums greatly decrease for those ages 45 64, falling by about $2,400, from $7,500 to $5,100 (Exhibit 2). People with incomes below 400 percent of the federal poverty level purchasing nongroup coverage in the exchange do not experience substantial changes in their average out-of-pocket spending, because federal premium subsidies protect them from price differences. However, average spending changes substantially for 294 Health Affairs February :2

6 Exhibit 2 Effects Of 3:1 Age Rating On Insurance Coverage And Spending, Compared With No Age Rating, :1 age rating No age rating All adults Nonelderly Nonelderly Nongroup premiums Single policyholders a $3,600 $4,200 $7,500 $5,200 b $5,000 $5,000 $5,100 $5,000 b All adults Individuals spending Average individual spending a 3,300 4,900 7,400 5,400 $4,800 3,600 5,100 6,400 5,400 $5,000 Below 200% of poverty 2,200 2,600 3,200 2,600 2,600 2,200 2,600 3,100 2,600 2, % of poverty 3,800 4,800 5,700 4,800 4,200 4,100 4,800 5,500 4,900 4, % of poverty or more 5,000 7,000 10,900 8,700 7,100 6,100 7,900 8,300 7,900 7,000 Coverage (millions of people) Employer coverage Exchange Non-exchange Individual coverage Exchange Non-exchange Public Uninsured SOURCE Urban Institute analysis, Health Insurance Policy Simulation Model, NOTE We simulated the provisions of the Affordable Care Act fully implemented in a Among adults with nongroup coverage only, rounded to the nearest hundred dollars. b Not applicable. those with higher incomes, increasing by about $1,100 for the youngest adults (from $5,000 to $6,100) and decreasing by $2,600 for the oldest (from $10,900 to $8,300). As a result of the premium changes, fewer young people and more people in the older age group purchase nongroup coverage, both inside and outside the exchange. The result is a net increase in nongroup coverage among the nonelderly by about 250,000, from 17.8 million to 18.1 million people. This net increase disguises a decline in nongroup coverage for children by about 380,000 and an increase in coverage for adults by about 630,000, from 15.4 million to 16 million people. Some children lose nongroup coverage because the premiums for their young-adult parents increase with the elimination of age rating. Ultimately, 27.2 million nonelderly people will be uninsured under a scenario with no age rating that is, 960,000 more than under the scenario with a 3:1 age rating. This occurs because eliminating age rating results in Medicaid enrollment decreasing from 70.1 million to 69.4 million (about 1 percent), predominantly among children. Because of the increases in premiums for younger adults, more will remain uninsured. Not interacting with the exchange, a greater number of these adults will not enroll their children in public coverage, either, particularly in the Children s Health Insurance Program. In addition, employer-sponsored insurance also declines as a result of eliminating age rating. As older employees find it more attractive to purchase coverage in the nongroup market without age rating, some employers will stop offering coverage. Federal premium subsidies are higher by about $1 billion without age rating because of the increased average premiums in the nongroup market, reflecting the compositional shift in the age of the nongroup exchange enrollees (see the online Appendix for details). 11,12 Employer spending on premiums falls slightly when age rating is eliminated because of the small decline in the number of people with employer-sponsored insurance. On balance, age rating reduces premiums for older adults and increases them for younger adults. Many younger adults are protected from the associated premium increases by incomerelated subsidies provided in the nongroup exchange. Relative to the standard case, the number of uninsured older adults declines by about one million; this decrease is roughly offset by increases in the number of uninsured adults in the two younger age groups. Eliminating The Employer Tax Credit We find that the employer tax credit has little effect February :2 Health Affairs 295

7 Exhibit 3 Enrollment In Small-Employer Coverage Under The Affordable Care Act, With And Without The Employer Tax Credit Enrollment (millions) Standard case Small group, exchange Small group, non-exchange No employer tax credit SOURCE Urban Institute analysis, Health Insurance Policy Simulation Model, NOTES We simulated the provisions of the Affordable Care Act fully implemented in For explanation of standard case, see text. Exhibit 4 on overall insurance coverage (see the online Appendix for details). 12 The employer tax credit is available to a subset of small employers purchasing coverage through the employersponsored insurance exchange. Exhibit 3 shows that eliminating the credit reduces the number of people purchasing coverage through the employer exchange by about 500,000. However, it increases the number of people obtaining employer-sponsored coverage through a small firm outside the exchange by about the same amount. Some small low-wage employers drop coverage if the credit is eliminated, while other employers begin to offer coverage.we find that the net effect of the employer tax credit is to move more employer coverage into the SHOP exchange. Relative to the standard case, premiums faced Employer-Sponsored Insurance Enrollment Under The Affordable Care Act, Based On Defined Size Of Small Employers Small firm, exchange 6.4% Small firm defined as up to 100 workers (N=154,596,000) Small firm, non-exchange 12.8% Large firm, employer coverage 80.8% Small firm defined as up to 50 workers (N=154,559,000) Small firm, non-exchange 8.2% Small firm, exchange 6.2% Large firm, employer coverage 85.7% SOURCE Urban Institute analysis, Health Insurance Policy Simulation Model, NOTE We simulated the provisions of the Affordable Care Act fully implemented in by small firms, premium subsidies, and spending by people in each income group are largely unaffected by eliminating the tax credit (see the online Appendix for details). 12 Defining Small Group Somewhat surprisingly, we find that changing the definition of small group from 100 or fewer workers to 50 or fewer workers has relatively little impact on the proportion of people covered through the exchange (Exhibit 4). This is largely because firms with employees can generally purchase coverage independently with lower administrative costs than the estimated administrative costs in the exchange. Administrative costs per enrollee decrease as employer size increases, because of the fixed nature of many of those costs. Thus, the purchasing efficiencies generated by the exchange will be largest for the smallest employers. Employers with workers are unlikely to achieve sizable administrative savings from the new exchanges. Consequently, for most firms of workers, there is little incentive to purchase coverage through the exchange when permitted. In addition, about 80 percent of workers employed in firms of fewer than 100 workers can also be categorized as being employed in firms of fewer than 50 workers. 13 As a result, the bulk of SHOP-eligible workers are unaffected by this policy difference. We estimate that the lower employer-size limit would reduce the number of people purchasing employer-sponsored insurance through the exchange by about 400,000. At the same time, about 400,000 more people would obtain coverage outside the exchange. There is almost no change in other forms of coverage, and the number of uninsured people is unchanged. There is little change in employer premiums, premiums in the nongroup market, or the cost of premium subsidies. There are also no meaningful differences in individuals spending in any income group or in total employer spending on premiums (see the online Appendix for details). 12 Conclusions Despite a great deal of uncertainty surrounding how states may set up their exchanges, results of these simulations suggest that variations in the particular policy choices and exchange designs studied will not dramatically alter the effects of the Affordable Care Act on overall costs and coverage. All of the effects noted here are relatively small. Therefore, these results suggest that states can make these design choices based on local support and preferences without fear of dramatic repercussions for overall coverage and 296 Health Affairs February :2

8 cost outcomes. For example, a state whose business community has strong concerns about the early expansion of the size of the small-group market to firms of up to 100 workers need not worry that delaying that expansion will significantly compromise the size of the SHOP exchange and SHOP s negotiating leverage with carriers. However, workers in those firms would experience a delay in receiving the full array of consumer protections provided by the Affordable Care Act. Certain choices, however, will cause noticeable distributional effects. Although reducing age rating bands to less than 3:1 will not have substantial consequences for overall coverage, the tighter bands will lead to fewer uninsured older adults and more uninsured younger adults, and they will differentially affect premiums for those purchasing coverage without the benefit of federal subsidies. Of all of the options we studied, merging the small-group and individual markets has the greatest positive impact on coverage and the greatest potential to lower premiums in the individual market. Lowering premiums in this market also lowers federal subsidy costs. States have other exchange design options available that are not analyzed here for example, designating exchanges as active purchasers, establishing roles for brokers and navigators, and standardizing benefit packages. These options may have stronger implications for coverage, costs, or both. The authors received funding from the Commonwealth Fund for this study. The authors acknowledge Habib Moody and Caitlin Carroll for their excellent research assistance. NOTES 1 Jost TS (Washington and Lee University, Lexington, VA). Health insurance exchanges and the Affordable Care Act: key policy issues. New York (NY): Commonwealth Fund; 2010 Jul. 2 Lueck S. States should structure insurance exchanges to minimize adverse selection. Washington (DC): Center on Budget and Policy Priorities; 2010 Aug. 3 Heberlein M, Brooks T, Artiga S, Stephens J. Holding steady, looking ahead: annual findings of a 50-state survey of eligibility rules, enrollment and renewal procedures, and cost sharing practices in Medicaid and CHIP, Washington (DC): Kaiser Commission on Medicaid and the Uninsured; Passel J, Taylor P. Unauthorized immigrants and their U.S.-born children. Washington (DC): Pew Hispanic Center; Blumberg L, Nichols LM, Banthin JS. Worker decisions to purchase health insurance. Int J Health Care Finance Econ. 2001;1(3 4): Nichols L, Blumberg L, Cooper P, Vistnes J. Employer decisions to offer health insurance: evidence from the MEPS-IC data. Paper presented at: American Economic Association Meeting; 2001 Jan. 5 7; New Orleans, LA. 7 Gruber J, Lettau M. How elastic is the firm s demand for health insurance? J Public Econ. 2004;88(7 8): Specifically, premiums are computed based on the costs of enrollees; employers choose whether to offer coverage at those premiums; and families decide what coverage, if any, to take. Premiums are then updated, and the cycle continues until decisions settle. 9 Kingsdale J. How small business health exchanges can offer value to their future customers and why they must. Health Aff (Millwood). 2012;31(2): For a technical explanation of the revenue provisions of the Reconciliation Act of 2010, as amended, in combination with the Patient Protection and Affordable Care Act, see Joint Committee on Taxation. Testimony of the staff of the Joint Committee on Taxation before the Subcommittee on Select Revenue Measures of the House Committee on Ways and Means Hearing on H.R. 3, the No Taxpayer Funding for Abortion Act, as reported by the House Committee on the Judiciary. Washington (DC): The Committee; 2011 Mar 16. Publication No. JCX Not only are premium subsidies higher as a result of eliminating age rating, but they are also distributed differently. Under a scenario with no age rating, people of different ages but similar incomes receive the same premium subsidy. In contrast, in a 3:1 age-rating scenario, older people face much higher premiums and would therefore receive larger subsidies than younger people with similar incomes. 12 To access the Appendix, click on the Appendix link in the box to the right of the article online. 13 Agency for Healthcare Research and Quality. Table 1.B. Number of private-sector employees by firm size and selected characteristics: United States Data from: Medical Expenditure Panel Survey: Insurance Component [Internet]. Rockville (MD): AHRQ; 2010 [cited 2012 Jan 31]. Available from: meps.ahrq.gov/mepsweb/ data_stats/summ_tables/insr/ national/series_1/2010/tib1.pdf February :2 Health Affairs 297

9 ABOUT THE AUTHORS: FREDRIC BLAVIN, LINDA J. BLUMBERG, MATTHEW BUETTGENS, JOHN HOLAHAN & STACEY MCMORROW Fredric Blavin is a research associate atthehealthpolicy In this month s Health Affairs, Fredric Blavin and coauthors from the Urban Institute s HealthPolicy Center discuss the results of their analysis of design options that states might choose in creating health insurance exchanges for individuals and small employers. Among the options examined with the use of the institute s Health Insurance Policy Simulation Model were creating separate versus merged small-group and individual markets, eliminating age rating, and restricting the small-group market to firms with 50 rather than 100 workers. The authors found relatively small differences in overall coverage levels and costs, although some options would result in higher premiums for some people while lowering costs for others. Blavin is a research associate at the Urban Institute s HealthPolicy Center, where he is working with the institute s model to estimate the cost and coverage implications of various state and national health insurance reform policies. Prior to joining the Urban Institute, Blavin was an economist at the Office of the National Coordinator for Health Information Technology, Department of Health and Human Services. Blavin earned his doctorate in health economics from the University of Pennsylvania. Linda J. Blumberg is an economist and senior fellow at the Health Policy Linda Blumberg is an economist and senior fellow at the Health Policy Center and a member of the Health Affairs editorial board. Her current work on health reform includes codirecting, with her coauthor and Urban Institute colleague John Holahan, a largescale, multiyear project monitoring and evaluating the implementation of the Affordable Care Act in ten states and nationwide. From August 1993 through October 1994, Blumberg served as a health policy adviser to the Clinton administration during its initial health care reform effort. Blumberg earned her doctorate in economics from the University of Michigan. Matthew Buettgens is a senior research associate at the Health Policy Matthew Buettgens is a senior research associate at the Health Policy Center. He is the lead analyst in the development and use of the Health Insurance Policy Simulation Model. Previously, at the Bureau of Labor Statistics, Buettgens wrote the data collection and analysis software for the Census of Fatal Occupational Injuries.Heearnedhisdoctoratein mathematics from the State University of New York at Buffalo. John Holahan is director of the Health Policy Holahan is director of the Health Policy Center. Much of his work over the past thirty years has focused on state health policy and issues of federalism and health. He has also published research on the reasons for the growth in the uninsured over the past decade as well as on the effects of proposals to expand health insurance coverage on the uninsured and on costs to federal and state governments. Holahan earned his doctorate in economics from Georgetown University. Stacey McMorrow is a research associate at the Health Policy Stacey McMorrow is a research associate in the Health Policy Center, where she leads a study on the effects of increased federal funding for community health centers on access to care for lowincome people. She has analyzed the potential impact of the Affordable Care Act on small employers, estimated savings from a variety of cost containment policies, and examined potential challenges to obtaining coverage for children under health reform. She earned her doctorate in health economics from the University of Pennsylvania. 298 Health Affairs February :2