By Anna Maria Barry-Jester and Ben Casselman
Nearly 9 million people gained insurance last year, a win for “Obamacare” as the president’s signature health care law expanded Medicaid and opened health insurance exchanges. And yet, 33 million Americans, 10.4 percent of the U.S. population, still went without health insurance for the entirety of 2014. Millions more were uninsured for at least part of the year.1 New data released this month shows they were disproportionately poor, black and Hispanic; 4.5 million of them were children.
It isn’t a surprise that some Americans still don’t have health insurance. Despite aiming to insure “everybody” in the U.S., the Affordable Care Act (ACA) left significant gaps in coverage, and decisions made by the law’s opponents have denied benefits to millions of people it was designed to help. But the new numbers reveal that most of the uninsured last year were people who should have been able to access insurance under the law. That presents a major challenge for President Obama in the final years of his term, but also an opportunity: Millions of Americans qualify for coverage but, for whatever combination of reasons, haven’t yet signed up.
The White House has said it will focus this year’s enrollment push on the remaining uninsured. So it’s worth taking a closer look at who those people were. Most of them — about 56 percent — fell into three major groups that were widely expected to have high uninsurance rates: immigrants, young adults and people in the so-called Medicaid gap. But that still left more than 14 million Americans who don’t have insurance and don’t fall into any of these categories.
7 million immigrants
Most uninsured noncitizen immigrants2 didn’t qualify for Medicaid, the government program that provides medical coverage for poor families, or the newly established health insurance subsidies that are meant to make insurance affordable to lower- and middle-income families. More than 4 million of them were most likely undocumented immigrants,3 who are explicitly barred from receiving Medicaid or buying insurance on the exchanges.
Another 600,000 immigrants had been in the country for less than five years and likely weren’t eligible for Medicaid and CHIP, a program for low-income children, though all income-eligible documented immigrants are entitled to subsidies on the exchanges. That left about 2 million who should in theory have qualified for Medicaid or subsidies on the exchanges (though many of them could have fallen in the Medicaid gap — more on that below). This group of longtime, legal residents has historically had lower rates of coverage than U.S. citizens; it will likely take time for them to make gains in coverage under the law.
One of every 5 immigrant children had no coverage last year, as well as 1 of every 10 senior citizens. Hispanics made up a majority of the 7 million uninsured immigrants, 79 percent, and Hispanic immigrants had an uninsured rate of 42.4 percent, nearly three times that of non-Hispanic immigrants. As we’ve written before, Hispanics disproportionately work low-paying jobs that don’t provide insurance, leaving them with fewer options than other groups. Immigrants also face a variety of challenges getting signed up for insurance on the marketplaces. More than 400,000 people were dropped from insurance this year because of missing or incomplete paperwork needed to prove eligible immigration status.
3.8 million in the Medicaid gap
Set aside the 7 million noncitizen immigrants, most of whom were never meant to be covered by the ACA, and 26 million uninsured remain. Of those, nearly 4 million were meant to qualify for insurance under the federal law but were later blocked from coverage. They fell into what’s known as the “Medicaid gap,” with incomes that were too high for Medicaid eligibility and too low to receive subsidies on the new health care exchanges.4
When the Affordable Care Act was written, it expanded the existing Medicaid program to cover everyone living below the poverty line, including childless adults who had previously been mostly excluded from the program. It also offered subsidies to people earning 100 percent to 400 percent of the poverty line (but not less) to buy private insurance on the exchanges. A Supreme Court decision left Medicaid expansion up to individual states, and about half chose not to expand the program. That left millions of low-income people ineligible for Medicaid but, counterintuitively, unable to qualify for subsidies on the insurance marketplaces because they earned too little.
This 3.8 million was a very poor group overall, but nearly half, 1.7 million people, lived in families whose incomes were less than 50 percent of the poverty line. Because Medicaid historically tended to cover just families with children, the bulk of those falling into the gap were working-age adults without children; nonetheless, about 800,000 parents with children living at home fell into the gap.
But it wasn’t just poor people who were more likely to be uninsured in states that didn’t expand Medicaid.5 At every income level, people in non-expansion states were more likely to be uninsured:
Even people earning six-figure salaries were more likely to be uninsured if they lived in a state that didn’t expand Medicaid. That may be because there was significant crossover between those states and the states that have federally run health insurance exchanges, which were notoriously problematic in their first year. Additionally, many of these states passed laws restricting where information about the exchanges could be shared, or which people could educate the public about the health care law.
7.7 million young adults
Without immigrants and people in the Medicaid gap, the total number of uninsured is about 22 million people, more than a third of whom are young adults ages 19 to 34. Unlike people in the previous two categories, these people weren’t excluded from the law’s benefits. In fact, these “young invincibles” were coveted: Because young people are generally healthier than older adults, they are cheaper to insure. Insurers need lots of them to sign up for coverage in order to keep premiums affordable for everyone else. However, because they are early in their working years — often unemployed or in jobs that don’t offer insurance — they have historically been the age group with the lowest coverage rate.
Early predictions that the young and healthy wouldn’t sign up and insurance premium costs would skyrocket didn’t come true (though not as many young adults signed up as was hoped). Uninsured rates decreased for all young adults by more than 5 percent from 2013 to 2014, leaving 17.1 percent of all 19- to 25-year-olds uninsured, as well as 18.2 percent of all 26- to 34-year-olds. But if we exclude young adult immigrants and poor residents of non-expansion states, those numbers fall dramatically, dropping the uninsured rate to 13.4 percent for 19- to 26-year-olds, whom the ACA allowed to stay on a parent’s insurance plan, and 8.2 percent for 26- to 34-year-olds.
The numbers suggest that young men, who might be less likely to seek regular medical care than women, didn’t always see insurance as worth the cost. Poorer young adults had higher uninsured rates; they might have qualified for subsidies, but the penalty for not having coverage last year was just $95 (or 1 percent of income, whatever was higher) and might have seemed like a better option for people without health problems who were earning low wages. In pre-ACA surveys, a small percentage of uninsured young adults said they didn’t need coverage, but a much larger number indicated that it was just too expensive.
And for a lot of young adults, insurance may have been unaffordable. Even excluding those living in “Medicaid gap” states, 1.5 million uninsured young adults were making less than $25,000 a year, giving that group an uninsured rate of 22.3 percent.
Unsurprisingly, uninsurance rates were particularly high for those who weren’t working. But even among those who did have jobs last year, rates were high compared with other age groups, likely because young adults often work jobs that don’t offer coverage.
The 14.4 million others …
That leaves a lot of people who weren’t caught up in the politics of immigration and Medicaid expansion, and weren’t the young adults so often discussed during the rollout of the ACA.
About a quarter of the 14.4 million remaining uninsured were children. Nearly all the rest were working-age adults; just 2.8 percent were 65 or older, the minimum age to qualify for Medicare. Nearly three-quarters of those working-age adults had jobs at least part of the year; nearly half of them worked full time all year long.
Some of the people in this category were very poor. More than 4 million theoretically qualified for Medicaid, 3.5 million in states that expanded coverage and half a million more in those that didn’t, and yet aren’t enrolled (the latter number is much smaller than the former because people in the Medicaid gap are already excluded).6 Nearly 3 million lived in households earning less than $25,000 a year.
But the group as a whole was economically diverse. Just less than half, 7 million, earned less than $50,000 a year — it’s likely the “affordable” options on the exchanges didn’t feel particularly affordable for this group. Another 3 million earned $50,000 to $75,000, at or above the U.S. median household income of $54,000 but still low enough that it might have made affording insurance a challenge, especially for those who didn’t qualify for subsidies. But that leaves the final, and perhaps least obvious, group: the 4.5 million who lived in households that earned more than $75,000 per year but remained uninsured. Many of these people were middle-aged Americans in relatively good health who may, like some of the young invincibles, have decided insurance wasn’t worth the cost.
But a surprising number, more than a million, were children. Most states offer Medicaid for kids in families earning incomes of up to 250 percent (or even more) of the federal poverty line, but a handful have a lower cutoff and have higher uninsured rates as a result. Others may have fallen into the “family glitch.” If one family member is offered insurance by an employer that is considered “affordable” for a single person, it doesn’t matter if it’s not affordable when purchasing for an entire family — they don’t qualify for marketplace subsidies.
It’s hard to say why these 14 million people weren’t insured, but the administration will have to figure that out if it wants to come close to the universal coverage the law intended.
The government measures health insurance coverage in various ways. The 33 million figure comes from the Current Population Survey, which asks respondents whether they had health coverage at any time in the previous year, in this case 2014. Other government surveys, such as the American Community Survey and the National Health Interview Survey, ask about coverage over shorter periods and show that a higher percentage of people were uninsured at least part of the year. For example, the NHIS found that 15.2 percent of the population under age 65 was uninsured during the first quarter of 2014. ^
Naturalized citizens qualify for all the benefits available to native-born citizens. ^
The census does not ask directly about immigration status, but it is possible to guess which survey respondents are likely to be undocumented based on their age, industry, ethnicity and other factors. Our estimate is based on a methodology developed by the Kaiser Family Foundation, which in turn based its approach on work by the State Health Access Data Assistance Center and by researchers at Penn State’s Population Research Institute. Kaiser explained its methodology in detail here. (The foundation kindly shared the R code behind its calculations, which we used with some modifications.) ^
The Current Population Survey doesn’t allow for a precise calculation of the Medicaid gap. Among other challenges, the CPS uses a different, broader definition of “household” than the one used in determining eligibility for Medicaid or the insurance exchanges (which, to make matters more complex, also differ from one another). For example, the census definition of “household” includes roommates, whose incomes would not be lumped together to determine health insurance eligibility.
In addition to households, the CPS also includes information on “families,” but that definition is narrower than the ones used by Medicaid and the exchanges. Unmarried partners, for example, aren’t in the same family as far as the census is concerned, but they are for insurance purposes.
The Kaiser Family Foundation has developed a methodology for reconciling the various definitions of household in order to more accurately estimate who falls into the coverage gap. A detailed explanation of that methodology can be found here. Kaiser kindly shared with us its R code, which we used in this analysis; any errors are our own.
Despite these adjustments, however, our numbers should be taken as a rough estimate. They cannot precisely identify how households are composed for health insurance purposes. In addition, the income figures in the CPS are self-reported and may contain inaccuracies. ^
For our analysis, we defined Medicaid expansion states as those that had expanded the program by Jan. 1, 2014, as well as Wisconsin, which didn’t expand under the ACA but covers people earning less than 100 percent of the federal poverty line. Michigan and New Hampshire were excluded from the analysis; both states expanded the program later in 2014. ^
Plus about 200,000 in Michigan and New Hampshire, which expanded Medicaid partway through the year. For the same reason discussed above, our analysis doesn’t allow us to identify precisely who would qualify for Medicaid. Some people who our analysis shows “should” qualify for Medicaid are likely misidentified and don’t qualify in the real world. But it seems likely that hundreds of thousands of people, at a minimum, aren’t getting Medicaid coverage they theoretically should.