CDC confirms what (nearly) everyone knew: The ACA has caused the uninsured rate to plummet.

The Centers for Disease Control has released their big National Health Interview Survey, which confirms what every other study, survey and poll has: The Affordable Care Act did indeed caused the uninsured rate to plummet last year. On the one hand, this isn't exactly surprising news to anyone other than former U.S. Senator/NH Governor Judd Gregg and current U.S. Representative Gary Palmer, but whatever.

On the other hand, this particular survey carries more weight in some ways than the previous ones from Gallup, the Urban Institute, the RAND Corporation or the Commonwealth Fund, since a) it's the official government survey and b) the survey pool is massive:

NHIS is a multistage probability sample survey of the civilian noninstitutionalized population of the United States and is the source of data for this report. The survey is conducted continuously throughout the year by NCHS through an agreement with the U.S. Census Bureau.

NHIS is a comprehensive health survey that can be used to relate health insurance coverage to health outcomes and health care utilization. It has a low item nonresponse rate (about 1%) for the health insurance questions. Because NHIS is conducted throughout the year— yielding a nationally representative sample each month—data can be analyzed monthly or quarterly to monitor health insurance coverage trends.

The fundamental structure of the current NHIS oversamples Hispanic, black, and Asian populations. Visit the NCHS website at: http://www.cdc.gov/nchs/nhis.htmfor more information on the design, content, and use of NHIS.

The data for this report are derived from the Family Core component of the 2009–2014 NHIS, which collects information on all family members in each household. Data analyses for the 2014 NHIS were based on 111,682 persons in the Family Core.

Due to the large amount of data/demographic breakdowns in the survey itself, you can get very different, but equally valid ledes from different news stories about the survey. For instance, here's Danielle Burger at Bloomberg Business:

The share of working-age people without health insurance fell by more than 4 percentage points in 2014, the biggest drop since the U.S. Centers for Disease Control and Prevention began reporting the data in 1997.

Last year, 16.3 percent of adults under age 65, or about 31.7 million people, lacked medical coverage, according to a CDC survey published Tuesday. That’s down from 20.4 percent a year earlier.

Meanwhile, here's Jeffrey Young at Huffington Post:

Millions of people gained health insurance last year as Affordable Care Act benefits took effect, according to the first official accounting by the federal government.

In 2014, 36 million U.S. residents, or 11.5 percent of the population, were uninsured on the day they were surveyed, a decline of 8.8 million people and 2.9 percentage points from the year before, according to the Centers for Disease Control and Prevention's National Health Interview Survey.

 

So, what's going on here? Well, take a look at those descriptors: The Bloomberg story refers to working-age people (ie, adults between 18-64 only) while the Huffington Post story includes everyone (children under 18 and seniors over 64, both of whom have always had lower uninsured rates anyway).

Again, there's nothing wrong with either story; it's just as important to know what's going on with working-age adults as it is with children and seniors...as long as you keep that in mind when reading the story.

In addition, it's important to keep in mind is that the numbers and percentages in the CDC's NHIS will look somewhat different from many of the other surveys, for some very important reasons:

• The other surveys tend to look at snapshot in time results for that particular date or quarter. For instance, the most recent Gallup and Urban Institute surveys were as of March 2015. The NHIS looks at the full 2014 calendar year.

• Related to this, the NHIS stops at the end of 2014, and therefore doesn't include any results from 2015...which means none of the several million additional people who enrolled via the ACA exchanges, additional Medicaid expansion in Pennsylvania and other states (Michigan topped out at 600K this spring) and so on.

• As referred to above, the other surveys tend not to include people over 65 and/or not to include children under 18. Those are kind of important, since those groups make up about 15% and 23% of the population respectively (38% combined, or around 122 million combined). The NHIS provides estimates of the uninsured rate for several different subsets of the total population, and as shown above, this means that different news stories may focus on different subsets, which is no doubt causing much confusion to many people this morning:
  • For everyone (all ages): 11.5% uninsured, or 36.0 million people. This is down from 16.0% in 2010 (when the ACA was signed into law) and 14.4% in 2013 (just prior to the ACA exchanges and Medicaid expansion kicking in).
  • For those under 65: 13.3% uninsured, down from 18.2% in 2010 and 16.6% in 2013
  • For children (those under 18): 5.5% uninsured, down from 7.8% in 2010 and 6.5% in 2013

There's a whole bunch of other important/useful stuff in the survey as well, of course, including tracking how many people were previously uninsured for part of the year or for more than a year, breakouts by region and state, and so forth.

In short, the overall news is excellent (if completely unsurprising given the massive amount of supporting data from Gallup/etc), but make sure to read the actual methodology and notes in both the survey itself and the news story about it before jumping to any conclusions.

As an aside, I can't believe that I have to point this out again, but given the fact that former U.S. Senator and New Hampshire Governor Judd Gregg doesn't seem capable of understanding the distinction, I guess I'll throw this in here as well: There's a huge difference between PERCENT and PERCENTAGE POINT.

• If a data point drops from 50% to 25%, that's a drop of 50 percent.
• If a data point drops from 50% to 25%, that's a drop of 25 percentage points.

Or, to put it in terms closer to the actual recent Gallup survey which tripped up Gregg:

• If the uninsured rate drops from 18% to 12%, that's a drop of 33%.
• If the uninsured rate drops from 18% to 12%, that's a drop of 6 percentage points.