Premiums under the Senate Better Care Reconciliation Act

The Senate Better Care Reconciliation Act (BCRA) would make significant changes to the amounts that people pay for nongroup coverage and for the care they receive under the Affordable Care Act (ACA). The tables below provide estimates of how premiums after taking into account tax credits would change for people currently enrolled in the federal and state marketplaces.

Under current law, people with incomes between 100 percent and 400 percent of the federal poverty level are eligible for premium tax credits to help them pay the premium for nongroup coverage purchased through the federal or a state marketplace if they do not have access to other affordable coverage. People are responsible for paying a specified percent of their income (“required income percentage”) toward the cost of the benchmark plan (the second-lowest cost silver plan in their area), and the federal government pays the remainder of the premium to their insurer; this amount is the person’s premium tax credit. The required income percentages people are responsible to pay vary with income: In 2017, people with incomes between 100 percent and 133 percent of poverty contribute 2.04 percent of income, while people with incomes between 300 percent and 400 percent of poverty contribute 9.69 percent of their income.¹ Because premiums vary with age but the share of income people are responsible to pay does not, older people receive larger premium tax credits than younger people with the same income but pay the same amount for the benchmark plan.

Beginning in 2020, the BCRA would make several significant revisions that affect the premium tax credits that people receive when they purchase nongroup coverage. First, the bill would revise income eligibility for premium tax credits, extending eligibility to people with incomes below poverty but capping eligibility at 350 percent of poverty. Second, the bill amends the way that premium tax credits are calculated so that the required income percentages vary with age and with income. Our estimates of the required income percentages under current law and the BCRA for 2020 are shown in the Appendix. The result is that on average people at younger ages would pay a lower share of their income to purchase a benchmark plan than they today while people at older ages would pay a higher share. Third, the bill reduces the value of the benchmark plans that are used to determine premium tax credits. The result is that a person who used their premium tax credit to purchase a benchmark plan would get a plan that on average would pay 58 percent of expected covered costs (a bronze plan), compared to 70 percent (a silver plan) under current law. A plan paying 58 percent of expected covered costs would have much higher cost sharing (e.g., deductibles) than a plan covering 70 percent of costs. This change is particularly important because the BCRA also would eliminate the cost sharing subsidies available under current law that reduce cost sharing and out-of-pocket limits for marketplace enrollees with incomes at or below 250 of poverty.

The bill also authorizes states to change the amount that premiums for adults can vary due to age, from 3:1 under current law to 5:1 (or a different ratio at state discretion). This would lower premiums for younger adults and raise them for older adults in states that made the change.²

Results

We estimated the average premiums that current marketplace enrollees would pay, after receiving any premium tax credit, for a benchmark silver plan in 2020 under current law and under the BCRA. Most current marketplace enrollees purchase silver plans, so we used those as the basis for a comparison of how much people would pay for equivalent coverage under the ACA versus the BCRA. The methods we used in making our estimates are described in more detail below.

Overall, marketplace enrollees would pay on average 74 percent more towards the premium for a benchmark silver plan in 2020 under the BCRA than under current law (Table 1). Younger enrollees would see modest increases on average (10 percent for those under age 18; 17 percent for those ages 18 to 34), while average premiums would more than double for enrollees ages 55 to 64. State-level results are in Appendix Table 2.

Table 1: Monthly Premium for a Silver Plan Among Exchange Enrollees (By Age), 2020
Age	ACA Premium After Tax Credit	BCRA Premium After Tax Credit	% Change
Under 18	$110	$120	10%
18-34	$145	$169	17%
35-44	$194	$271	39%
45-54	$208	$403	94%
55-64	$271	$583	115%
65 and Older	$310	$660	113%
Overall (All Ages)	$197	$342	74%
Source: Kaiser Family Foundation

These results vary significantly by income as well (Table 2). Marketplace enrollees with incomes below 200 percent of poverty would see an average increase in their premium costs of 177 percent, while higher income enrollees would see an increase of 57 percent.

Table 2: Monthly Premium for a Silver Plan Among Exchange Enrollees (By Income and Age), 2020
	Income Below 200% of Poverty			Income 200% of Poverty or Above
Age	ACA Premium After Tax Credit	BCRA Premium After Tax Credit	% Change	ACA Premium After Tax Credit	BCRA Premium After Tax Credit	% Change
< 18	$26	$58	121%	$176	$170	-4%
18-34	$57	$103	82%	$247	$247	0%
35-44	$69	$149	117%	$296	$369	25%
45-54	$67	$215	223%	$323	$556	72%
55-64	$69	$272	294%	$399	$782	96%
65 +	$76	$296	288%	$439	$862	96%
Overall	$61	$168	177%	$311	$489	57%
Source: Kaiser Family Foundation

There are important differences by age within these income groups: among enrollees with incomes below 200 percent of poverty, those in 18 to 34 age group would see an average increase of 82 percent while those in the 55 to 64 age group would see an average increase of 288 percent. Among enrollees with incomes 200 percent of poverty and above, enrollees in the 18 to 34 age group would not see an increase while those age 55 to 64 would see their premium costs almost double.

Discussion

The vast majority of marketplace enrollees would pay higher premiums in 2020 for a silver plan. Older and lower income enrollees see the biggest increases. These results are driven by several provisions in the BCRA. First, the BCRA reduces the value of the benchmark plan used to calculate the premium tax credits (from a plan that, on average, pays 70 percent of expected costs to a plan that pays 58 percent of expected costs). Lowering the benchmark means that marketplace enrollees could enroll in what is roughly a bronze plan by paying their required income percentage, but that they would need to pay the entire difference in premium to enroll in the silver level plans that are most prevalent today. The second factor is the change in the required income percentages under the BCRA, which generally would reduce what younger adults would be required to pay but increases the amounts paid by older adults, particularly those at higher incomes. Among people with higher incomes, reducing the maximum income eligibility for premium tax credits from 400 percent of poverty to 350 percent of poverty increases costs for some marketplace enrollees, particularly people at higher ages who face relatively high premiums. Increasing the permitted premium variation due to age also would increase premiums for older adults not eligible for premium tax credits.

These significant increases in the costs for silver plans may cause some or many marketplace enrollees to look to lower-value bronze-level plans, which they could purchase by paying their required income percentage. For younger marketplace enrollees, they generally would pay less under the BCRA to purchase a bronze level plan than they would pay for a silver plan under current law; older enrollees, however, generally would pay more for a bronze level plan under the BCRA than they would pay for a silver plan under current law. Moving down to bronze level plans, however, would expose enrollees to much higher cost sharing than in silver plans, and for many enrollees who now receive cost-sharing subsidies, the increases would be very large. The BCRA would eliminate the cost sharing subsidies provided under current law beginning in 2020.

The reduction in the value of the benchmark plan, along with the elimination of cost sharing subsidies, raises questions about whether lower income people would continue their coverage under the BCRA. While premiums after premium tax credits might be somewhat lower for younger enrollees purchasing bronze plans, their cost sharing would likely be thousands of dollars higher; the average deductible for bronze plans in 2017 with a combined deductible for medical and prescription expenses is $6,105; this compares to an average deductible of $809 for plans with cost sharing reductions for people with incomes between 150 and 200 percent of poverty and $255 for people with incomes between 100 and 150 percent of poverty. Many people with low incomes would have a difficult time paying the cost sharing under the benchmark plans in the BCRA, and may decide they do not want to pay even a relatively small premium for a plan that they would struggle to use.

Because of the short time between the release of the discussion draft and the planned debate and vote in the Senate, we were unable to address all of the provisions that might affect premiums under the BCRA. Our analysis assumes that unsubsidized premiums for a 40-year-old would ultimately be the same under the ACA and BCRA. In the score released today (June 26), the Congressional Budget Office expects the difference in unsubsidized premiums for a 40-year-old between the ACA and BCRA to be quite small in 2026, which is consistent with our assumption; however, they expect BCRA premiums to be relatively lower in 2020, due in part to larger federal funding available in 2020 to reduce premiums. While these changes would have some impact on our results, the impact would be muted because we have focused on the amount that people pay after tax credits, and for most marketplace enrollees, those amounts are determined by their required income percentage and not the actual plan premium. For these people, the actual premium affects the amount of their tax credit, but not what they would pay for a benchmark plan. Generally lower premiums would affect our results primarily for those marketplace enrollees who would pay the full premium with no premium tax credit under the BCRA. These generally would be people with higher incomes or younger people facing very low premiums such that the full premium would be less than their required income percentage.

Methods

We used data from the March 2016 Current Population Survey, the 2016 National Health Interview Survey and administrative data about the income and demographic distribution of the population enrolled in the federal and state marketplaces to construct a model of nongroup enrollees.

To impute marketplace enrollment status for each individual reporting directly purchased private health insurance to the March 2016 Current Population Survey, we applied a series of modeling techniques to the health insurance units (HIUs) described here in order to model the division of individuals holding nongroup coverage between those enrolled in a marketplace and those enrolled outside of a marketplace. Using the same multiply-imputed technique described here, we repeated our draw of each state’s nongroup population ten times to accurately account for sampling error.

We revised our Uninsured Calibration described in here to more closely align with the insurance coverage movements shown by the recent CDC publication of full-year 2016 National Health Interview Survey (NHIS) estimates.³ This CDC document shows continued gains in public coverage during the year and a leveling-off of private insurance coverage gains after the early 2016 Marketplace enrollment surge that mirrors administrative data sources. Our previous publications on this topic were calibrated to NHIS 2016 first quarter estimates; however, both HHS-published effectuated enrollment in the Exchanges at the end of March 2016 and also the insurer rate filings used to estimate the size of the off-marketplace population more closely align with the trends exhibited by NHIS 2016 full-year statistics.⁴ Calibrating to national CDC estimates allowed for on- and off-marketplace sampling targets consistent with administrative enrollment at the state level.

For each state’s on-marketplace and off-marketpace nongroup population, we drew purchasing units across five strata, each informed by federal data. For each state and the District of Columbia, we sampled subsidy-eligible marketplace enrollees both above and below 250% FPL, followed by a small group of ACA subsidy-eligibles forgoing help in the off-exchange market.⁵ Consistent with this administrative data, we also sampled a small group of wealthier nongroup enrollees (those not eligible for subsidies) into the Exchanges, and moved the remaining nongroup individuals into the off-exchange market. For non-immigrants with incomes below 100% of poverty, we calculate the amount of their required premium contribution as though their income were 138% of poverty. To most accurately reflect the age and income distribution of the Exchanges, each marketplace-purchasing unit received a sampling probability proportional to the average monthly subsidy per person within the state. At the conclusion of these ten repeated state sample draws, our average advanced premium tax credit (APTC) per month landed at $284 nationwide (compared to the $291 reported by HHS) and within $30 of the actual amounts displayed on table two of the HHS effectuated enrollment report for every geography except for the state of Connecticut.⁶ This close match of estimated APTC dollars for forty-nine states and the District of Columbia reflected a high degree of accuracy of the demographic (primarily age and income) distribution of our sampled exchange population.

To compare the effect of the Senate’s proposed Better Care Reconciliation Act (BCRA) against current law under the Affordable Care Act (ACA), we attached both the second lowest cost silver and the lowest cost bronze plan premiums to each individual in each local market.  These 2017 premiums from the Kaiser Family Foundation’s Subsidy Calculator matched CPS respondents at the state and metropolitan area-level, with smaller areas not disclosed by the U.S. Census Bureau computed using a population-weighted average premium across the aggregation of non-metro areas.  Matching our prior eligibility analyses, we computed ACA eligibility and subsidy receipt using the second lowest cost silver plan available to the HIU as the benchmark plan.  To reflect the 58% Actuarial Value (AV) level stated by the BCRA, we used each geographic area’s lowest cost bronze plan as the benchmark plan for each HIU.  In a small number of geographies without a bronze plan option for purchase on the 2017 exchanges, we used 85% of the silver plan as that local area’s benchmark plan premium.  For each individual’s premium calculations under the BCRA, we relaxed the ACA’s 3:1 age rating to a 5:1 age band in all states that did not have community rating requirements in place prior to 2014.  We followed CBO and HHS inflation factors to project all dollar values and thresholds to calendar year 2020.  Using CBO’s economic projections,⁷ we inflated the 2015 income amounts in the 2016 CPS for each HIU to 2020 dollars.  We increased premium dollars from 2017 to 2020 and both ACA and BCRA premium caps from 2014 to 2020 using Centers for Medicare & Medicaid Services Office of the Actuary projections.⁸ Our analysis assumes that states do not take up a waiver under the BCRA.