find a plan

Does the IRS change how much I’ll have to pay for my health insurance each year?

Image: Andrey Popov /

Q. I’ve seen that the percentage of income I have to pay for my health insurance seems to change each year, based on IRS guidance. Does this change the actual amount that I pay for my coverage each year? And how has the American Rescue Plan (ARP) affected this?

A. The IRS does make an annual change — usually quite small — to the percentage of income that you have to pay for self-purchased (individual/family) health coverage. But there’s a lot more to it than just the percentage of income that the IRS says you have to pay for the benchmark plan (which has been significantly reduced for 2021 and 2022 as a result of the American Rescue Plan). It also depends on how your income changes relative to the federal poverty level (FPL, which changes every year), how your health plan’s premium changes, whether you make a change to your coverage during open enrollment, and the fact that you continue to get older each year.

Let’s take a look at how it all works, both in terms of the normal rules and the temporary changes that the ARP imposed in order to help the country recover from the COVID pandemic:

The ACA has a lot of moving parts. Various aspects of the regulations have to be updated annually, including the affordability rules. Initially, the IRS laid out guidelines detailing the percentage of tax filers’ income that they would be expected to contribute towards their own premiums, assuming their income doesn’t exceed 400% of the poverty level. (For premium subsidy purposes, income means an ACA-specific version of modified adjusted gross income, and subsidy eligibility is based on how that income compares with the prior year’s FPL.)

(Note that although there is usually an income cap of 400% of the poverty level in order to qualify for premium subsidies, that has been temporarily removed for 2021 and 2022 under the ARP. For those two years, nobody purchasing coverage in the marketplace has to pay more than 8.5% of their income for the benchmark plan, regardless of their income level.)

Applicable percentages increased each year from 2015 through 2017

Then in July 2014, the IRS released Revenue Procedure 2014-37, in which they explained the changes to the percentage of income that subsidy recipients would have to pay (known as the applicable percentage) if they selected the second-lowest-cost Silver plan in the exchange in 2015.

A few months later, in November 2014, the IRS published Revenue Procedure 2014-62, which laid out the changes to the applicable percentage for 2016. And in 2016, they published Revenue Procedure 2016-24, which detailed the changes for 2017.

For 2015, 2016, and again for 2017, there was a slight increase in the applicable percentage numbers. Because the applicable percentage climbed each year, there was often an assumption that everyone who gets subsidies was paying more for their health insurance (after subsidies) in each successive year. But that’s not necessarily the case, as we’ll see in a minute.

Applicable percentages decreased for 2018

For 2018, the applicable percentages decreased slightly when compared with 2017, meaning that the percentage of income that people had to pay (after subsidies) for their coverage was slightly lower at all income levels than it was in 2017. (See Revenue Procedure 2017-36 for 2018 numbers.)

This means that before accounting for other factors (including age and income changes), people buying coverage for 2018 had to pay a slightly smaller portion of their income for the second-lowest-cost Silver plan than they paid in 2017.

In 2018, people who were eligible for premium subsidies had to pay between 2.01% and 9.56% of their income for the benchmark plan in their area. Of course, people don’t have to pick the benchmark plan – they can pick a higher-cost plan and pay more, or they can pick a lower-cost plan and pay less.

Applicable Percentages increased for 2019 to the highest they’ve ever been

For 2019, the applicable percentages went back up again, and were the highest they had ever been. Even though the applicable percentages increased again for 2021 (before being drastically lowered by the ARP), they were still lower than they were in 2019.

But as described below, people who were subsidy-eligible and whose income hadn’t increased since 2014 were paying less in after-subsidy premiums in 2019 than they were paying in 2014 (assuming they selected the benchmark plan in both years), due to the annual growth in the poverty level since 2014.

The details for 2019’s applicable percentages are in Revenue Procedure 2018-34. For 2019 coverage, people who were eligible for premium subsidies paid between 2.08% and 9.86% of their income for the second-lowest-cost silver plan, after subsidies.

Applicable percentages decreased again for 2020

For 2020, the applicable percentages decreased again. The details are in Revenue Procedure 2019-29, which was published in July 2019. For 2020 coverage, people who were eligible for premium subsidies paid between 2.06% and 9.78% of their income, after the subsidy was applied, for the second-lowest-cost silver plan (ie, the benchmark plan).

Applicable percentages increased for 2021, but the ARP then sharply reduced them

When 2021 began, the applicable percentages were higher than they had been in 2020, meaning that at each income level, people would have to pay a slightly larger percentage of their income for the benchmark plan. But Section 9661 the American Rescue Plan made some very significant reductions to the applicable percentage table for 2021 and 2022.

(And even before the applicable percentages were reduced by the ARP, it’s important to remember that the poverty level also increased for 2021, meaning that you would have had to get a raise in order to remain at the same percentage of the poverty level; the 2020 poverty level numbers are used for 2021 subsidy eligibility determinations. If you don’t get a raise, your income as a percentage of the poverty level would be lower, resulting in a smaller amount that you’d have to pay for your health coverage after the subsidy is applied.)

The applicable percentages for 2021 were initially posted in Revenue Procedure 2020-36, requiring subsidy-eligible enrollees to pay between 2.07% and 9.83% of their income for the benchmark plan, with no limit on how high premiums could be if a household’s income was over 400% of the poverty level. But the American Rescue plan has reduced this range. It’s now 0% to 8.5%, with no income cap for subsidy eligibility. (A subsidy is not available if the benchmark plan would cost less than 8.5% of your income without a subsidy, so people with very high incomes will still not qualify for a subsidy, as their coverage would not exceed 8.5% of their income. But subsidy eligibility can now extend well above 400% of the poverty level for some enrollees, particularly those who are older and live in areas where health insurance is particularly expensive.)

For 2021 coverage, under the modified rules implemented by the American Rescue Plan, subsidy-eligible enrollees who buy a plan in the exchange have to pay the following percentages of their income, after the subsidy is applied, for the benchmark plan:

  • Income up to 150% of poverty = 0% (ie, the subsidy is enough to make the benchmark plan premium-free)
  • 150% to 200% of poverty = 0% to 2%
  • 200% to 250% of poverty  = 2% to 4%
  • 250% to 300% of poverty = 4% to 6%
  • 300% to 400% of poverty = 6% to 8.5%
  • 400% of poverty or higher = 8.5%

All of these percentages are retroactive back to the start of 2021, even though the subsidy amounts won’t start to show up on until April 1, 2021, and won’t be automatically adjusted for current enrollees (they’ll have to log back into the exchange to resubmit their enrollment and get the new subsidy amount to update in real-time). Any additional premium tax credits can be claimed when marketplace enrollees file their 2021 tax returns.

Making sense of applicable percentages

Those numbers might make your eyes glaze over. But the following examples will show how they actually affect premiums from one year to the next. The first set of examples show how Bob’s premiums changed from 2014 to 2021, if his income increases each year to keep pace with increases in the federal poverty level. Below that, you’ll see what happens if Bob’s income hasn’t increased since 2014. And in both cases, we’ll show how the American Rescue Plan reduces the amount that Bob has to pay for his health coverage in 2021, making coverage more affordable than it’s been in any of the previous years. Yes, there’s math involved, but never fear — it’s pretty straightforward.

Example: Bob’s premium changes from 2014 to 2021, if he gets annual raises that keep his income at 200% of the poverty level

Bob’s MAGI (modified adjusted gross income as calculated for the ACA’s premium tax credits) is equal to 200% of poverty. In this example, we’re going to assume he’s received a pay raise each year in order to keep his income at 200% of the poverty level. In 2014, his applicable percentage was 6.3%, for 2018, it was 6.34%, for 2019, it was 6.54%, for 2020 it was 6.49%. For 2021, it was initially 6.52%, but the ARP has reduced it to just 2%. That means he only has to pay 2% of his income for the benchmark plan instead of 6.52%, and his subsidy amount will be larger to make up the difference.

For subsidy purposes, poverty level determinations are based on the year during which open enrollment begins. Since the open enrollment period for 2021 coverage took place in 2020, the government uses 2020 poverty level guidelines for determining subsidy-eligibility for any plans that have 2021 effective dates. (poverty level numbers are published by HHS in January each year, but that’s after open enrollment for that year’s coverage has already ended).

2014: $1,448 in premiums, with an income of $22,980

If Bob was earning 200% of poverty level when he got his 2014 plan, his MAGI was $22,980 (based on the 2013 poverty level) and his applicable percentage (the amount he had to pay for the second-lowest-cost silver plan) was 6.3 percent. So he had to pay $1,448 in annual premiums in 2014 ($22,980 x 0.063). His subsidy paid the rest of the premium, assuming he selected the second-lowest-cost silver plan (ie, the benchmark plan).

2018: $1,529 in premiums, with an income of $24,120

But for 2018, if Bob was still earning 200% of poverty level, his MAGI had increased to $24,120, since the poverty level has increased. His applicable percentage was 6.34%, which equaled $1,529 in annual premiums ($24,120 x 0.0634). That’s about $81 more in annual premiums than he had to pay in 2014. But in order to maintain his percentage of poverty level at 200%, he was earning $1,140 additional dollars per year.

So while his net premium had increased, his income was also increasing (and note that this is comparing his premiums in 2014 with his premiums in 2018; if we just look at 2017 versus 2018, the slight decrease in the applicable percentage for 2018 nearly exactly offsets the increase in the poverty level, making his net premium just about exactly the same in 2018 as it was in 2017).

2019: $1,588 in premiums, with an income of $24,280

For 2019, the applicable percentage increased again, and the poverty level also increased. (2018 poverty levels were higher than they were for 2017.) If Bob was still earning 200% of the poverty level when he signed up for 2019 coverage, it meant his income had grown to $24,280. He had to pay 6.54% of that for the second-lowest-cost Silver plan, which amounted to $1,588 in annual premiums ($24,280 x 0.0654). So Bob was paying $140 more in annual premiums in 2019 than he was paying in 2014 — but his income was $1,300 higher than it was in 2014, since we’re assuming he’s stayed at 200% of the poverty level.

2020: $1,621 in premiums, with an income of $24,980

For 2020, the applicable percentage decreased, but the poverty level continued to increase. To keep Bob at 200% of the poverty level, his income had to grow to $24,980 for 2020 (based on 2019 poverty level numbers, since that’s what’s used for the 2020 coverage year). He had to pay 6.49% of his income for the benchmark Silver plan, which amounted to $1,621 in annual net premiums. That’s $33 more in annual premiums than he paid in 2018, and $173 more than he paid in 2014. But his income in 2019 was $2,000 higher than it was in 2014.

2021: $1,664 in premiums, with an income of $25,520

For 2021 coverage, the poverty level is higher than it was for 2020 coverage. But although the applicable percentage was initially higher as well, it’s been drastically reduced by the American Rescue Plan. In Bob’s case, with an income of 200% of the poverty level, his new applicable percentage in 2021 is only 2%. For Bob’s income to stay at 200% of the poverty level, it had to increase to $25,520 in 2021. But instead of having to pay $1,664 in after-subsidy premiums for the benchmark silver plan (the amount he would have paid under the normal applicable percentage for 2021), he only has to pay $510, which is 2% of his income.

This is a total annual savings of about $1,154 over what he was paying before the American Rescue Act, and it’s also significantly lower than the amount he’s had to pay for his coverage in any year since the ACA was implemented.

What if Bob doesn’t get a raise?

But what if he doesn’t get a raise, and his MAGI is still $22,980 in 2021? That means his income will be 180% of the poverty level, instead of 200%, so his applicable percentage will be less than the 2% that would have applied if his income had grown to keep pace with the increases in the poverty level ($22,980 divided by $12,760 is 1.8; that means Bob’s income is 180% of the 2020 federal poverty level, which we use to calculate 2021 subsidy eligibility).

The calculation

To calculate applicable percentages for incomes that are somewhere within each range on the chart, you can use the formula that’s explained in CFR 1.36B-3. (Scroll down to just underneath the applicable percentage chart, and look at example 2.) In the case of Bob, it looks like this:

Part 1

180 – 150 = 30
200 – 150 = 50
30/50 = 0.6

Basically, you look to see what income range you’re in. (Bob is between 150 and 200% of poverty range.) Then you just figure out how far along the income range you are.  In this case, Bob’s income is 60% of the way along the range that goes from 150% to 200% of poverty.

Part 2

2 – 0 = 2
2 x 0.6 = 1.2

0 + 1.2 = 1.2

The second part of the calculation is to look at the applicable percentage range for 2021 coverage that corresponds to Bob’s income range (0% to 2% — meaning that the percentage of income he’ll have to pay for the second-lowest-cost silver plan is somewhere between those two percentages). And then you just figure out what number is 60% of the way along that applicable percentage range. In Bob’s case, it’s 1.2 percent.

His applicable percentage for 2021 is 1.2% (note that it would have been 5.57% without the American Rescue Plan), and his net premium, after his subsidy is applied, is just $276 in annual premiums (without the American Rescue Plan it would bave been $1,280; this was less than it was in 2014, but still much higher than it is now that the ARP has been enacted).

In 2020, with the same $22,980 income, Bob would have been at 183% of the poverty level and his applicable percentage was 5.68%. So his net premium for the benchmark silver plan was $1,305 in 2020 — lower than it was in 2014, 2018, and 2019. And as we saw above, it’s much lower in 2021 due to the ARP. But even before the ARP changed the ruled, Bob’s net premium would still have been lower in 2021 than it was in 2020, despite the fact that the applicable percentages increased across all income levels. Despite the increase in the applicable percentage, Bob would still have seen a decrease in the total dollar amount that he had to pay for the benchmark silver plan in 2021. Because Bob’s income hasn’t kept pace with the poverty level, the amount he has to pay in net premiums is declining over time (and as we’ve seen in these exampled, the ARP has reduced Bob’s after-subsidy premium to much lower levels than it’s ever been before).

Applicable percentages increased for 2019 and again for 2021, but so did the poverty level – and you have to consider them together

There are a lot of moving parts here. Although the applicable percentages for 2019 were the highest they’ve been since this system was implemented (and initially increased from 2020 to 2021, albeit not quite to 2019 levels) the poverty level has continued to increase each year. So people whose incomes have not increased in several years could have found that they were paying less in total premiums in 2021 than they were paying a few years earlier, even before the American Rescue Plan boosted subsidies across the board.

How the applicable percentage is calculated — it’s changed a bit in recent years

The general idea behind the adjustment to the applicable percentage table is to keep up with changes in premium growth as they relate to changes in income. If health care costs increase faster than income, we all have to pay a larger chunk of our income for health care. But if the economy does well and the ACA’s efforts to curb healthcare spending are successful, it’s also possible for the applicable percentage to decrease — as was the case for 2018 and for 2020.

The formula for the adjustment to applicable percentage is just premium growth since 2013 divided by income growth since 2013. But the methodology for calculating each of those numbers has changed over time.

Premium growth used to be based on average per-enrollee premiums for employer-sponsored plans, in terms of how much those premiums had changed since 2013. But for 2020, HHS finalized a methodology change that incorporates premium changes in the individual market, as well as premium changes for employer-sponsored plans.

This was widely expected to result in an increase in applicable percentages for 2020, but when the numbers were published in July 2019, the applicable percentages for 2020 were lower than they had been for 2019 (without the methodology change, applicable percentages would likely have decreased even more for 2020, as the estimation was that they would be 2.7 percent higher in 2020 with the new calculations that incorporate premium changes in the individual market).

Income growth was based on changes in GDP per capita for plan years 2014 through 2016, but HHS finalized a new formula that has been used to calculate income growth since 2017. The new formula calculates income growth based on per-capita personal income (PI) rather than per-capita GDP. The two methods would likely generate similar numbers, but HHS considers per-capita PI changes to be a more accurate reflection of how per-capita income changes from one year to the next.

The IRS also added a provision that allows for an additional adjustment for years after 2018 to reflect the premium growth rate relative to the growth in the consumer price index [the additional adjustment is described in §36B(b)(3)(A)(ii)(II)]. But the next section in that code [§36B(b)(3)(A)(ii)(III)] says that the additional adjustment is only needed if the total amount the government spent on premium subsidies and cost-sharing reductions in the previous year was more than 0.504% of the previous year’s gross domestic product. For 2019, 2020, and again for 2021, the IRS determined that the additional adjustment was not necessary.

Since subsidies are also a function of the poverty level — which generally adjusts upward each year — there’s a built-in factor that essentially ensures that people who are impacted by a higher applicable percentage are also enjoying at least a modest increase in income that outweighs the additional premiums.

Average benchmark premiums dropped in 2019, 2020, and 2021, resulting in smaller subsidies. But the American Rescue Plan is dramatically increasing subsidies for 2021 and 2022

Premium subsidy amounts are based on the relationship between an applicant’s income and the federal poverty level, but they’re also highly dependent on the premium that the applicant would have to pay for the benchmark plan. The formula described above is how subsidy amounts are determined in every state (keeping in mind that Alaska and Hawaii have their own poverty level numbers).

But what if you don’t buy the benchmark plan? Many areas have dozens of plans available for sale in the exchange, and only one of them is the benchmark plan (it changes from year to year, but it’s always the second-lowest-cost silver plan). If you buy a plan that’s not the benchmark plan and you’re eligible for a subsidy, you still get the same subsidy amount that you’d have received if you had purchased the benchmark plan, but it’s applied to the price of the plan you select instead.

Nationwide, overall average premiums in the individual market increased slightly in 2019, decreased slightly in 2020, and increased slightly in 2021 (overall, premiums are much more stable than they were in 2017 and 2018, when they grew rapidly in most areas). But average benchmark premiums decreased in all three years. The decrease in benchmark premiums happened in some areas because an existing insurer lowered their rates, but in other areas it’s because a new insurer joined the market and began offering lower-priced silver plans than the ones that were already available.

Since premium subsidy amounts are based on the cost of the benchmark plan’s premium in each area, it’s not surprising that premium subsidy amounts have been dropping over the last few years. Across all enrollees who are receiving premium subsidies, the average subsidy amount was $550/month in 2018, dropped to $539/month in 2019, and dropped again, to about $492/month in 2020.

But as we saw above, the whole point of premium subsidies is to cover the difference between the actual cost of the benchmark plan and the amount the enrollee is expected to pay for that plan based on a specific percentage of their income. So even if the amount the enrollee is expected to pay remains unchanged, the premium subsidy amount will go down if the full-price cost of the benchmark plan goes down.

The best course of action is to actively shop for your coverage each year during open enrollment. Never let your plan automatically renew without checking first to make sure that it’s still the best option. And if you become eligible for a special enrollment period mid-year due to a  qualifying event, make sure you actively compare all of the plans available to you before making a decision about whether to keep your existing coverage or make a change.

Louise Norris is an individual health insurance broker who has been writing about health insurance and health reform since 2006. She has written dozens of opinions and educational pieces about the Affordable Care Act for Her state health exchange updates are regularly cited by media who cover health reform and by other health insurance experts.

Find affordable health plans.

Helping millions of Americans since 1994.

(Step 1 of 2)

Related articles

0 0 votes
Article Rating
Notify of
Inline Feedbacks
View all comments
Would love your thoughts, please comment.x