On October 5, 2021, we corrected several errors in the “Changes in Spending” section and in Table 5 of this brief arising from a coding error that did not apply all cost-sharing reductions to household spending. These errors affect estimates of household and total health spending among nonexpansion states in 2022 under each of the policy options examined:
In Table 5, we also combined estimates of spending on premium tax credits, cost-sharing reductions, and reinsurance into a single estimate, labeled "Marketplace subsidies and reinsurance," for the federal government and state governments. A previous version of the table contained separate estimates for each of these three spending categories. The total of the three spending components is the same as reported earlier, but small amounts reported as cost-sharing reductions should have been attributed to premium tax credits.
The Affordable Care Act (ACA) allows states the option of expanding Medicaid coverage for people under age 65 to 138 percent of the federal poverty level (FPL) ($17,774 for an individual and $36,570 for a family of four in 2021). 1 As of this writing, 12 states have neither expanded Medicaid nor passed ballot initiatives to expand Medicaid as allowed by the ACA. We project that there will be 14.6 million uninsured, 2 of whom 5.8 million are below 100 percent of FPL in these 12 states in 2022. 3
Lawfully present individuals in these states with incomes above 100 percent of FPL ($12,880 for an individual and $26,500 for a family of four in 2021) can use premium tax credits and cost-sharing reductions to obtain ACA marketplace coverage, provided no family member has an offer of coverage deemed affordable under the law. If a person or household has an income below 100 percent of FPL, they are not eligible for marketplace subsidies. 4 In states that have not expanded Medicaid under the ACA, few in this group are eligible for Medicaid under current rules. 5 This leaves millions of people uninsured and without access to low-cost health insurance.
Because of the very large gap between traditional Medicaid eligibility levels and 100 percent of FPL, a large number of uninsured adults could gain coverage if the gap were filled. For example, Texas covers parents below 17 percent of FPL and Alabama 21 percent; childless adults are generally not covered in nonexpansion states. Moreover, there are some people with incomes between 100 percent and 138 percent of FPL who are ineligible for marketplace subsidies because family members have an employer offer of coverage for the worker that is deemed affordable under the ACA. Since there is no such employment-related test for Medicaid eligibility, Medicaid expansion makes more people in this income range eligible for affordable coverage.
Absent Medicaid expansion, the federal government could extend eligibility for premium tax credits and cost-sharing reductions to people below 100 percent of FPL, thereby covering the large majority of this group, at 100 percent federal cost. To avoid giving current expansion states a financial incentive to reverse their decisions, however, the federal government could increase the matching rate — the Federal Medical Assistance Percentage (FMAP) — from 90 percent to 100 percent for the ACA Medicaid expansion population in those states. This, of course, adds to the cost of each option. We show separately the cost of increasing the FMAP matching rate to 100 percent in current expansion states.
Subsidy estimates in this analysis are for nonexpansion states only, as they are necessary to make the coverage expansion work. Increasing subsidies would affect all states and would cost considerably more than shown here.
In this brief, we examine the coverage and cost impacts of three alternative approaches to filling the gap:
Table 1 shows premium and cost-sharing subsidies under each of the three options. The first column shows these subsidies before the ARPA’s enactment. This schedule would be reinstated after 2022 without subsequent legislation to make the ARPA enhanced tax credits permanent. 7
Option 1 has the same premium and cost-sharing subsidies that existed prior to the ARPA but extends subsidy eligibility to people with incomes below 100 percent of FPL. Note that, unlike Medicaid, there are premiums of just over 2 percent of income, with disenrollment for nonpayment. 8 Also, coverage is at 94 percent of actuarial value — that is, plans cover 94 percent, on average, of the costs of covered health benefits — while Medicaid coverage generally has zero or nominal cost sharing. 9
Option 2 also extends eligibility to people with incomes below 100 percent of FPL but enhances premium subsidies. As shown in Table 1, premium subsidies are considerably more generous at each income level and extend above 400 percent of FPL ($51,520 for an individual and $106,000 for a family of four in 2021). Premiums are eliminated for those below 150 percent of FPL. This option not only extends assistance paying for coverage to those below 100 percent of FPL but also improves affordability for those throughout the income distribution.
Option 3 extends eligibility to people with incomes below 100 percent of FPL but increases both premium and cost-sharing subsidies. Subsidies are tied to the gold plan level. Option 3 also adds a federal reinsurance plan. The main difference between options 2 and 3 is the improvement in cost-sharing subsidies and reinsurance.
Medicaid expansion would mean no premiums and cost sharing below 100 percent of FPL. Premium and cost-sharing subsidy schedules in the marketplaces would remain the same as they were prior to the ARPA.
Exhibit 1 and Table 2 show the coverage effects in 2022 for each option, emphasizing the changes relative to current tax credit eligibility with pre-ARPA premium tax credits.
Option 1 extends coverage to 3.0 million people on net. About 3.4 million would receive marketplace coverage, but 363,000 would drop their employer coverage because they now have a more affordable option. The number of people enrolled in Medicaid increases slightly, because when more people seek marketplace coverage, more family members eligible for Medicaid or the Children’s Health Insurance Program (CHIP) get enrolled. The result is a reduction in the number of uninsured people by 3.0 million, including 200,000 people who leave nongroup coverage that does not comply with ACA benefit requirements, such as short-term health plans.
Option 2, which extends ACA coverage to people with incomes below 100 percent of FPL and also expands premium subsidies throughout the income distribution, covers 4.6 million more individuals in nonexpansion states than under current law with pre-ARPA subsidies. Enrollment in private nongroup coverage grows by 4.9 million people, nearly all of whom have newly subsidized marketplace coverage. Enrollment in employer coverage drops by 379,000, as individuals use the enhanced subsidies to move to nongroup coverage.
Medicaid enrollment increases for the same reason as in Option 1 but is higher because the increase in marketplace enrollment is much larger. The number of uninsured falls by 4.6 million people, including 269,000 who leave ACA-noncompliant nongroup coverage. Option 2 reduces the number of uninsured considerably, because it enhances subsidies for those with incomes above 138 percent of FPL as well as extends them to those below 100 percent of FPL.
Option 3, which further expands both premium and cost-sharing subsidies, covers an additional 5.0 million previously uninsured individuals in nonexpansion states. Enrollment in private nongroup coverage expands by 5.5 million people, nearly all of whom have newly subsidized marketplace coverage. The number of people with employer coverage drops by 558,000. Medicaid coverage increases by a small amount. The number of uninsured falls by 5.0 million, including 364,000 who leave noncompliant, nongroup coverage. Option 3 reduces the number of people who are uninsured more than the other two options because of the more generous premium and cost-sharing subsidies.
Medicaid expansion increases the number of people with coverage by 4.1 million, with total Medicaid enrollment expanding by 6.2 million, though not all new enrollees would have been uninsured previously. In particular, there is a drop of 1.2 million in private nongroup coverage, mostly because of people with incomes between 100 percent and 138 percent of FPL switching to Medicaid from marketplace coverage. About 828,000 people would switch from employer coverage to Medicaid. The result is a drop in the uninsured of 4.1 million, including 173,000 who leave noncompliant nongroup coverage.
Medicaid expansion thus does not cover as many people as Options 2 and 3 (enhanced subsidies), because there is no change in affordability above 138 percent of FPL. Nearly all changes in coverage are for people with incomes below 138 percent of FPL.
Table 3 shows the changes in nongroup and Medicaid coverage by income.
Option 1 extends marketplace coverage to people earning less than the poverty level but does not improve premium and cost-sharing subsidies. It brings new nongroup or Medicaid coverage to an additional 3.1 million individuals with incomes below 138 percent of FPL. There are small increases in coverage at higher incomes, because the expansion of coverage for people below the poverty level positively affects the nongroup market risk pools, causing a small reduction in premiums that attract some additional enrollees.
Option 2 extends new nongroup or Medicaid coverage to 3.5 million people below 138 percent of FPL in nonexpansion states. This reflects the extension of coverage to people below the poverty level as well as the improvement in premium subsidies. The improvement in subsidies at higher income levels results in an increase in new nongroup or Medicaid coverage of 221,000 between 138 percent and 200 percent of FPL, 809,000 between 200 percent and 400 percent of FPL, and 475,000 above 400 percent of FPL. This option has large coverage gains both because of the expansion of subsidized coverage below the poverty level and the increased generosity of subsidies above 138 percent of FPL.
Option 3 extends new nongroup or Medicaid coverage to 3.5 million people below 138 percent of FPL in nonexpansion states. The improvement in premiums and cost-sharing subsidies results in an increase in coverage of 270,000 between 138 percent and 200 percent of FPL, 1.3 million between 200 percent and 400 percent of FPL and 510,000 above 400 percent of FPL. This option results in the largest coverage gains, much of which is above 138 percent of FPL.
Medicaid expansion increases Medicaid or nongroup coverage by 4.9 million below 138 percent of FPL. Unlike marketplace premium tax credits, Medicaid does not make families ineligible for assistance if family members have offers of other coverage deemed affordable; thus, of the three options, it offers the largest enrollment gain for the lowest income group. 10 Medicaid has lower cost sharing than the other options and no or low premiums. There are also potential differences in benefits. The overall expansion of enrollment in Medicaid is 6.2 million, but there is a decline in nongroup and employer coverage, leaving 5.0 million more people at all income levels with Medicaid or nongroup coverage. Individuals move from private nongroup and employer coverage into a much more affordable and cost-protective Medicaid plan.
Prior Medicaid expansions have led to a modest increase in enrollment among those already eligible. These results show slightly higher Medicaid and CHIP coverage above 138 percent of FPL for the same reason.
Exhibit 2 and Table 4 show the reduction in the uninsured (including the insured who leave noncompliant plans) by income level.
Option 1 opens the ACA marketplace to people below the poverty level and reduces the number of uninsured below 138 percent of FPL by 2.9 million. There are small reductions in the uninsured above 138 percent of FPL because of reductions in premiums as the risk pool improves. Overall, the number of uninsured falls by 3.0 million.
Option 2 is like option 1 but with enhanced premium subsidies. It reduces the number of uninsured by 3.2 million among people with incomes below 138 percent of FPL in nonexpansion states. There are also sizable reductions in the number of uninsured at higher income levels. Overall, the number of uninsured falls by 4.6 million.
Option 3 has both enhanced premium and cost-sharing subsidies and reduces the number of uninsured below 138 percent of FPL by 3.3 million in nonexpansion states. Overall, the number of uninsured falls by 5.0 million, more than in option 2 because the cost-sharing subsidies are more generous. Both options 2 and 3 have dual objectives: filling the gap and improving subsidies.
Medicaid expansion reduces the number of uninsured below 138 percent of FPL by 4.1 million. This option increases coverage the most for low-income adults. There is a slight reduction in the uninsured above 138 percent of FPL. Overall, the number of uninsured people falls by 4.1 million.
Exhibit 3 and Table 5 show changes in spending for each option for 2022. Exhibit 4 shows changes in spending for each option over 10 years (2022–2031).
Option 1, which extends coverage to people below the poverty level, results in an increase in federal spending of $15.1 billion in 2022 ($181 billion over 10 years). Federal spending on Medicaid and marketplace subsidies rises by $16.9 billion.
Federal spending on uncompensated care declines by $1.8 billion. Following assumptions made in similar analyses by the Congressional Budget Office (CBO), we assume that 50 percent (or $0.9 billion) of uncompensated care savings will be realized as Medicare Disproportionate Share Hospital (DSH) savings; how the remainder of the reduction in the uncompensated care is realized is uncertain.
If the federal matching rate in expansion states were increased to 100 percent of FPL, federal spending would increase by another $12.1 billion; overall federal sending would increase by $27.2 billion, or $327 billion over 10 years.
Household spending increases slightly ($1.4 billion) as households spend less on premiums, because of the coverage expansions, but spend more out of pocket. State governments also see reduced demand for uncompensated care of $1.1 billion overall. Employers reduce spending on premiums by $2.7 billion because of the small reduction in employer-sponsored insurance. Total health spending in these states increases by $11.1 billion.
Option 2 increases spending by the federal government in nonexpansion states by $22.5 billion in 2022 (the corresponding 10-year estimate is $270 billion). The greater marketplace premium and cost-sharing subsidies, plus a small increase in Medicaid spending, amount to $26.1 billion. There is a reduction in federal spending on uncompensated care of $3.7 billion, because fewer people are uninsured. Of this, Medicare DSH payments fall by $1.8 billion; the distribution of the rest is unknown.
If the federal matching rate in expansion states were increased to 100 percent of FPL, federal spending would increase by $12.1 billion. Under that scenario, federal spending increases by $34.6 billion, or $415 over 10 years.
Households also spend slightly less than under current tax credit eligibility with pre-ARPA tax credits: $3.7 billion less on premiums and $0.1 billion less in total out-of-pocket spending, because more people are purchasing coverage. States would see lower demand for uncompensated care, with costs falling by $2.3 billion. Employers reduce spending on premiums by $2.9 billion, owing to the small decrease in employer coverage. Total health spending increases by $10.4 billion as a result of the coverage expansion and expanded subsidies.
Option 3 increases federal spending in nonexpansion states by $27.9 billion, or $335 billion over 10 years. Increased marketplace subsidies amount to $28.5 billion. Expanded reinsurance adds $3.3 billion. There is a reduction in uncompensated care of $3.9 billion; Medicare DSH payments fall by $2.0 billion, half of the overall reduction in federal uncompensated care spending. If federal matching rates are increased in expansion states, federal spending increases by $12.1 billion, for a increase of $40 billion, or $480 billion over 10 years. Households spend $6.0 billion less as more people buy health insurance. Employers reduce spending on premiums by $4.0 billion. Total health spending in nonexpansion states increases by $12.2 billion.
With Medicaid expansion, the federal government increases spending by $30.5 billion in 2022, or $366 billion over 10 years. Medicaid spending increases by $43.1 billion but there is a reduction in marketplace subsidies ($9.7 billion) and uncompensated care ($2.8 billion) that offsets some of this. While states see $5.2 billion in new Medicaid spending because of required matching payments — offset somewhat by lower spending on uncompensated care — this does not take into account important savings to state governments. Most states have found that Medicaid expansion has a negligible impact on state budgets or even results in net savings. 11
There are savings to households of $4.0 billion, primarily reduced out-of-pocket costs. Employers reduce spending on premiums by $3.9 billion because low-wage employees enter Medicaid rather than keeping their employer coverage. Overall health care spending increases by $23.6 billion.
The biggest increase in health coverage for low-income people comes through Medicaid expansion. This offers a broad benefit package and has no or nominal cost sharing. As a result, it costs the federal government more to cover low-income people than under the other two options, which have greater cost sharing and premiums. Options 2 and 3 cover more people at higher incomes by generously subsidizing premiums and cost sharing. Those subsidies are, however, generally much lower than the cost of offering those people full Medicaid coverage.
The results presented above reflect the effects of joint policies of expanding subsidies and filling the Medicaid coverage gap. An alternative way to analyze this issue is to estimate the cost of filling the gap for those below 100 percent of FPL, building on an existing policy that has previously expanded subsidies. In option 1 above, federal spending for people in the Medicaid gap increases by $16.6 billion as they become eligible for subsidies. However, spending on health grows somewhat less — $15.1 billion — largely because the cost of premium tax credits falls for people already in the subsidized nongroup market. Premium tax credits are smaller, because the new enrollees have lower expected health costs than the currently insured population, and those lower costs are spread among all participants in the nongroup market, lowering premiums across the board. These lower premiums lead to a small increase in enrollment among those not eligible for premium tax credits.
In the second and third options, federal spending for people in the gap increases by more ($17.2 billion and $18.1 billion, respectively), since the subsidies for the new enrollees are more generous than in the first option and slightly more people enroll. The increase in overall federal health spending in these options ($22.5 billion in option 2 and $27.9 in option 3) is greater than the additional spending for people in the gap for two reasons. First, the more generous subsidies apply also to people outside the gap (those with incomes above 100 percent of FPL), who are eligible for assistance with nongroup premiums. As they do under option 1, premiums fall as more people enter the market, but because subsidies are greater, the government pays more of the remaining costs (and affordability for households improves). Second, the more generous subsidies increase enrollment significantly for those above 100 percent of FPL, which increases the total cost of subsidies.
As shown in Exhibit 5, the 10-year costs of just filling the gap, building on existing subsidies, are $199 billion, $207 billion, and $217 billion, respectively. The cost of increasing the federal match in expansion states increases these amounts to $344 billion, $352 billion, and $362 billion, respectively.