THE MAIN POINT - People without Health Insurance

Current Population Survey - CPS - 2005 & 2006


There are not 47 million who want but who cannot get health insurance. Neither the Census Bureau's Current Population Survey (CPS) nor the CDC's National Health Interview Survey (NHIS) show this.


The number of people without health insurance is not the same as the number of people who are unable to get insurance.

The number of people without health insurance reported by the major surveys includes EVERYONE who does not have health insurance - and everyone includes millions of people who could get but have CHOSEN not to buy health insurance or who are unaware of government programs AND all of their children.

The surveys do not report how many people want but cannot get insurance. That number must be estimated from their reported numbers and other available statistics.


To estimate the number of people unable to get health insurance, one has to remove wealthy people who don't want it, young adults who decide they do not need it, people temporarily in between jobs who decide not to buy insurance before a new job, and all those who are unaware of government programs that they could use.

Our estimates indicate that the number of people who want health insurance but are unable to get it is less than half the over 40 million in the surveys. The true number of people who cannot get health insurance is probably in the 20's of millions or even less.


For a full discussion and details on how we performed this estimation, see the People without Health Insurance Model.


The reason why we believe that the true number of people who want but cannot get health insurance is smaller is based on estimates we have made from numbers in this survey summary and others.

Here are numbers consistent with our assertion. Please review comments with the NHIS survey for a more complete picture.

These numbers suggest that a large fraction of the uninsured are dominated by people with some combination of being a young adult and/or having a moderate to high income. These are people who could, if they chose, acquire health insurance.

Nonetheless, there are still a lot of people who probably have difficulty getting insurance. The number is just not over 40 million people.

Read more below to understand how to interpret the data.

Sources of the Data

Tables 1-7 come from the Current Population Survey for the years 2005 and 2006 prepared by the Bureau of the Census. The total uninsured reported for 2006 is 47.0 million and for 2005 is 44.8 million.

There is another study performed by the National Health Interview Survey (NHIS) which we have also summarized. Both of these studies are surveys not censuses and are subject to sampling error and error in method.

Questions Answered - Not Answered by the Numbers

The uninsured numbers in the following CPS tables answer the following question: "How many people have been uninsured for any reason AND uninsured for the entire previous year?"

The numbers DO NOT answer the question, "How many people are uninsured and want but cannot get health insurance?"

Neither survey answers the first question. In fact, the CPS survey includes:

We are trying to estimate the number of people who cannot get insurance, using both the CPS and NHIS numbers and believe that number to be between 25% and 50% of the 47 million in our tables or between 10 and 20 million people.

The CPS and NHIS Numbers Answer Very Different Questions

The Census' Current Population Survey (CPS) which reports 44.8 million uninsured in 2005 and 47.0 million uninsured in 2006) answer the question, "How many people have been uninsured for any reason AND uninsured for the entire previous year?"

These numbers would therefore not include the NHIS component of people insured for less than one year (about 9.1 million), yet the Census number is bigger, 44.8 million compared to 41.6 million. One would expect that the CPS should be smaller since it is not supposed to include people uninsured for less than one year and the NHIS does.

Removing the 9.1 million who have been uninsured for less than 1 year from the NHIS to make the numbers more comparable leads to 44.8 (CPS) vs. 32.5 (NHIS), i.e. there is a very large discrepancy between these two surveys for 2005. These discepancies persist in 2006, 47.0 (Census) vs. 30.7 (NHIS).

This disrepancy has not gone unnoticed by the government agencies that publish the numbers. (See References). Some analyses claim that respondents to the CPS may be unable to recall how long they have been uninsured and not answer correctly. We question whether this is a full explanation. Certainly, people might forget if they have been uninsured for 11 vs. 12 months but are less likely to confuse 1 month with 12 months.

We think that there must be a difference in methodology or sampling that is leading to this difference but cannot be sure. Whatever the reason, it is certainly a questionable practice of choosing one number over the other when it is clear that at least two serious attempts have come up with very different numbers. For your interest, there are other estimates which we have not reported.

Uninsured Illegal Aliens

The surveyors have not attempted to include or exclude this group and we know of no way to estimate whether there is a fair sample of them in this survey.

Table 1. Number of Uninsured by Age

This table shows a comparison of uninsured for 2006 and 2005. Note that the number of children INCLUDES children of those categories that must be excluded to esimate the number of people who want but cannot get insurance.

In other words, the number of chronically uninsured children is less than this number and probably a lot less.

Return to table

Table 2. Number of Uninsured by Household Income

This breaks down the uninsured by household income. It does not appear to have the same unnaccounted for component in the NHIS numbers.

Return to table

Table 3. Number of Insured by Type

This table shows how people are being insured: privately by their employers, through government programs or through insurance that they purchase themselves. The number of people who purchase insurance on their own is about 9%.

Return to table

Tables 4 and 5. Number of Insured by Type and Age for 2005 and 2006

This table shows the number of people insured and ununinsured by type and age. Note the drop in number of people privately insured from the 45-54 to the 55-64 group. Since the percentages in tables 6 and 7 do not show such dramatic shifts, we think this may be due to a decline in the population due to death. We will try to confirm this.

Return to table

Tables 6 and 7. Percent of Insured by Type and Age for 2005 and 2006

This table illustrates, among other things, the dynamics of the young adult bracket. As people age, and presumably change both their employers and priorities, they are increasingly covered by some type of insurance

Return to table