Obamacare is not perfect. It does not provide everyone access to basic healthcare. It does not control costs. And it has had negative impacts on some. Even its most staunch proponents know there is work to be done. But the debate rages over whether what is needed is an improvement to Obamacare or a “repeal and replace.” In that context there is a highly disputed statistic; namely that “20 million people have health insurance coverage as a result of Obamacare.” Many on the right call it an outright lie. But what is the evidence for either side’s view? That is a statement of fact rather than opinion, so it should be objectively verifiable. So as part of my effort to cut through partisan rhetoric to find the reality, I looked. It is surprisingly difficult to arrive at an objective conclusion, which I suppose is not surprising at all in light of the competing rhetoric. But my conclusion: True. An important caveat though: By “insured” I mean those who are not “uninsured,” which is defined to exclude those covered under public programs.
Let’s start with the data. I asked a friend whom I know to espouse the “it’s a lie” view, what his data source is. He responded, “CBO,” referring to the Congressional Budget Office. So I started there. I found an interesting blog post on how the CBO estimates the number of uninsured.
CBO uses data from the MEPS—Household Component as a benchmark to estimate the number of people who are uninsured. No single source of data exists regarding people’s coverage status, and the only reliable information on uninsured people comes from federal surveys. Because of differences in design across surveys, CBO compares estimates from the MEPS—Household Component with estimates from the NHIS and the CPS. It considers the strength of each survey to arrive at a final estimate of the number of people who are uninsured.
Source: CBO.
Ok so right off the bat we have the CBO telling us that there is not any single source of data. Well, that complicates things. Then they tell us that federal surveys provide the only reliable information, but even those are not sufficiently reliable to rely on any one, so they compare three different ones. And what are they? MEPS? Here is a brief explanation with links for those wanting to read more:
- Medical Expenditure Survey – Household Component – MEPS is a survey by the Agency for Healthcare Research and Quality, or AHRQ, which is part of HHS. There looks to be quite a lot of data available online, though I have not explored it.
- NHIS – The National Health Interview Survey, or NHIS, is conducted by the National Center for Health Statistics, which is part of the CDC. So with typical government efficiency, it would appear that we have two agencies doing more or less the same survey. Perhaps there are variations that relfect each department’s mission. I have no idea. But it does appear that the NHIS data is the more widely cited. This is consistent with its assertion that “the National Health Interview Survey (NHIS) is the principal source of information on the health of the civilian noninstitutionalized population of the United States.” Here, too, there is much data available online.
- CPS – The Current Population Survey is a joint project of the Census Bureau and the Department of Labor. It is also cosponsored by the Bureau of Labor Statistics, whatever that means. As you might guess, its emphasis is on the workforce, but it does collect data on health insurance coverage as well. Its website provides links to download data, but it appears not as user-firendly.
Ok, so that’s what MEPS, NHIS and CPS are. What does CBO conclude from them? That I could not figure out. I could find nothing in which the CBO reported its own conclusions. So I asked my friend who relies on the CBO data. Surely he knew. He pointed me to the US Debt Clock website. Seriously? This is one of the problems with political discourse. So little of it is based on actual facts. Not that the fine purveyors of the US Debt Clock site are wrong. But they clearly have an agenda and are not the actual source of the data, so it is reasonable to question the accuracy of what they report. In fairness, I will say that when you scroll over the uninsured figure, a window reveals that accurately describes who are included in the federal survey totals as uninsured: persons not having any private insurance, Medicare, Medicaid, CHIP, or military plan.
Given that “20 million” is a round figure to start with, perhaps precision is not the goal here. So I picked one data set, NHIS, which, as mentioned, appears to be the leading source. Here is what I found. These are percentages of the US population uninsured:
| Year | Uninsured at the time if interview | Uninsured for at least part of the year | Uninsured for more than a year |
| 1997 | 15.4 | 19.5 | 10.4 |
| 2005 | 14.2 | 17.6 | 10.0 |
| 2010 | 16.0 | 19.8 | 11.7 |
| 2011 | 15.1 | 19.2 | 11.2 |
| 2012 | 14.7 | 18.6 | 11.1 |
| 2013 | 14.4 | 17.8 | 10.7 |
| 2014 | 11.5 | 16.5 | 8.4 |
| 2015 | 9.1 | 13.2 | 6.2 |
| 2016 (Q1) | 8.6 | 12.3 | 5.1 |
Source: NHIS.
Even when we look at a single source, there is no one definition of what it means to be uninsured. There are arguments for any of the three definitions. “At the time of the interview” provides a consistent snapshot in time and has the advantage of not being subject to memory failings. “At least part of the year” points to the problem of how fragile coverage is for many. And “more than one year” seems to reflect people the system is fully and consistently leaving behind. Looked at another way, though, it does not really matter to our present inquiry. We are looking at differentials. As it seems more relevant to the impact of Obamacare, I will compare the 2016 data (which only runs through March) and 2013. Under the three categories, the number of uninsured fell by 5.8% of the population, 5.5%, and 5.6%, respectively. So about 5.5% to 5.8% less of the US population is “uninsured.”Source: NHIS. (Apologies for the formatting.)
So how many people is 5.5% of the population? The answer is actually more complicated because the population changed over that same time period. According to the US Census Bureau, there were 316,204,908 people as of July 1, 2013, and 323,127,513 as of July 1, 2016. So let’s do a bit of math:
- 2013: 14.4% of 316,204,908 is right around 45.5 million people who were uninsured at the time of the interview.
- 2016: 8.6% of 323,127,513 is right around 27.8 million people who were uninsured at the time of the interview.
So this necessarily simplified approach gives us roughly 17.7 million fewer uninsured in 2016 than there were in 2013. But that does not tell the whole tale. Remember our question was not how many fewer are left uninsured (an important question, no doubt) but how many are newly insured. So we need to add to this figure the population increase in 2016 over 2013 of roughly 6.9 million people. So even though we have 6.9 million more people, we have 17.7 fewer uninsured. It follows then that there are about 24.6 million people who were uninsured in 2013 but who are insured in 2016.
This is, as noted, a highly simplified analysis. There are many other ways in which people benefitted from or were harmed by Obamacare. And it tells us nothing about the impact that high out of pocket costs had on people actually being able to use their insurance, the sustainability of Obamacare in its present form, or whether the newly insured are in the private insurance market or on Medicaid or CHIP. The Medicare/private distinction matters to some, so I will conclude with a brief look at that.
From the same NHIS report, here is data on the percentage breakdown of uninsured, public and private coverage. As an aside, note the sharp drop in private coverage before Obamacare.
| Year | Uninsured at the time if interview | Public health plan coverage | Private health insurance coverage |
| 1997 | 15.4 | 23.3 | 70.7 |
| 2005 | 14.2 | 26.4 | 67.3 |
| 2010 | 16.0 | 31.4 | 60.2 |
| 2011 | 15.1 | 32.4 | 60.1 |
| 2012 | 14.7 | 33.4 | 59.6 |
| 2013 | 14.4 | 33.8 | 59.5 |
| 2014 | 11.5 | 34.6 | 61.8 |
| 2015 | 9.1 | 35.6 | 63.2 |
| 2016 (Q1) | 8.6 | 36.2 | 63.8 |
Source: NHIS.
Applying the same math using the census figures from above:
- 2013
- 14.4% uninsured = 45.5 million
- 34.6% on public plan = 109.4 million
- 59.5% on private insurance = 188.1 million
- 2016
- 8.6% uninsured = 27.8 million
- 36.2% on public plan = 117.0 million
- 63.8% on private insurance = 206.2 million
This reflects an increase in public plan enrollment of 7.6 million people. I will assume that most of that is Medicaid/CHIP. Even so, there is other data that suggests the public component is higher. HHS reports that over this period Medicaid/CHIP enrollment has grown by 17 million people. (Source: HHS.) Perhaps I will go back and try to reconcile these figures at a later date, but I am out of time for now.
This post was updated to correct egregious math errors in the Medicaid section and to add the following:
After initially posting this, I engaged in a lengthy online debate with a friend who argued – without reference to this post – that nobody received coverage under Obamacare because Bernie says 29 million are uninsured now and Obama said in 2009 that over 30 million were uninsured then. My contention was that he was willfully ignoring the facts because he thought he had caught two politicians in a word trap. But, ok, if he wanted to rely on their words then he should at least be precise in quoting what they said. Obama’s was a carefully worded statement that omitted illegals and those who were eligible for Medicare but did not sign up. See the explanation here. There is nothing inconsistent about the two statements. But even if there were, Bernie is right, as backed by the NHIS data.