A Model to Account for the Uninsured
(without health insurance)

The purpose of the model below is to demonstrate that a set of reasonable assumptions using professional surveys can account for the total number of uninsured people in the US.

The conclusion is that most of the so-called "47 million without health insurance" are NOT "people who want but cannot get health insurance."

NOTE 1: ALL NUMBERS OF PEOPLE ARE IN THOUSANDS.

NOTE 2: This is a work in progress. We will continue to refine it and will appreciate any suggested changes, especially in the model assumptions.

.

Return to The Uninsured (CDC - NHIS)

Return to The Uninsured (Census - CPS)

The Model

Table 1. The Model - People who want but cannot get health insurance.

A Model NOT a Survey

The numbers above are estimated based on survey results and reasonable estimates of important factors. We are not suggesting that they are the actual numbers.

The purpose of the model below is to demonstrate that a set of reasonable assumptions using professional surveys can account for the total numbers of uninsured people in the US.

We are suggesting that any credible explanation of the total number of uninsured reported by either the CDC's NHIS (National Health Information Survey) or the Census Bureau's CPS (Current Population Survey) at least contain the categories of this model, and perhaps others.

Conclusions

It is not plausible to assume that either the CDC-NHIS total uninsured numbers or the Census Bureau CPS (Current Population Survey) for people without insurance for 12 months can be can be interpreted as "the number of people who want but cannot get insurance."

This model demonstrates that estimating the contribution of human decision making priorities and well-known processes in the economy have significant contributions to the number of people without health insurance at any given point in time.

A more precise model, like the one presented, is needed to interpret these results. Some of the numbers that we have estimated could be refined with data that the government's statisticians have ready access to.

Presentation Comments

We present our model first as a summary table (above) and support each row of the summary with a detailed table below. Our goal is to make our assumptions and estimation method completely transparent so that you can edit our model with your assumptions and sources.

In each of the detailed tables below, we started with a relevant population obtained from the National Health Information Survey (NHIS survey table 1) - or from the more detailed report in the references - #3. We then apply relevant factors to come up with our estimate.

Some of our consistency checks use the Census Bureau's Current Population Survey (CPS).

We show sources of data for each data element. If we have guessed the factor, we designate the source as "our assumption." If we have calculated a number from a combination of our assumptions and other sources, we designate it as "estimated."

In most cases, the estimated numbers can be computed by multiplying the source and assumption numbers above them.

The last table is our estimate of the number of children per adult in various age groups. We use these numbers to compute the number of children associated with each adult category of uninsured.

Wealthy People

Table 2. Wealthy People


Notes:

1. We estimate that the most relevant group of wealthy people will be those in families with incomes over $500,000. This fraction was obtained from IRS statistics (not a survey). Our own table used a $200,000 income cutoff so we went back to the original IRS data.

Young Adults

Table 3. Young Adults.


Notes:

1. We obtain the population by from the 18-44 number by assuming that there are the same number of people in each 1 year age range.

2. We think our assumption of only 10% of young people choosing not to buy insurance is quite conservative.

3. Singles factor - Since those choosing not to have insurance will be less likely to have children than the overall 18-30 age group, we further reduced our number to account for this. Multiplying the singles factor by the number of children/adult gives our esimate of the fraction of people in this group with children - ~5%.

People Between Jobs (18-44)

Table 4. People in between Jobs (18-44).


Notes:

1. Job turnover rate - Note that this number is a combination of the unemployment rate and those that are temporarily in-between jobs primarily by choice.

2. Not Electing COBRA - We assume only 20% of the people in between jobs choose not to get COBRA insurance from their employer or acquire it privately. In other words, we assume that most do insure in some way when they are in between jobs.

3. Consistency Comment - The sum of the adults without insurance in this table (18-44) and the next (45-64) - 4,940,000 - is consistent with NHIS table 5 that reports 9,107,000 people report that they have no insurance because they lost their job or changed employment. We have kept this number smaller because the loss of insurance is not temporary for the whole group.

People Between Jobs (45-64)

Table 5. People in between Jobs (45-64).


Notes:

1. Job turnover rate - Note that this number is a combination of the unemployment rate and those that are temporarily in-between jobs primarily by choice.

2. Not Electing COBRA - We assume only 15% of the people in between jobs choose not to get COBRA insurance from their employer or acquire it privately. In other words, we assume that most do insure in some way when they are in between jobs.

3. Consistency Comment - The sum of the adults without insurance in this table (18-44) and the previous (18-44) - 4,940,000 - is consistent with NHIS table 5 that reports 9,107,000 people report that they have no insurance because they lost their job or changed employment. We have kept this number smaller because the loss of insurance is not temporary for the whole group.

People Unaware of Their Eligibility for Government Insurance

Table 6. People Unaware of Government Insurance


Notes:

1. Base Populations - To get a base population for this category we assumed that all would come from families with incomes less than $35,000. Not all NHIS respondents reported their income so we had to estimate an additional amount to take care of the unaccounted for people.

2. Fraction with private insurance & Medicaid - In the general population, the fraction of people with private (non-government) ranges from ~66% to ~75% depending on age. We assumed that in this group it would be much smaller at 20?. Overall, the uninsurance assumption (30%) times the total people in the category (76.6 million) gives 23 million which is consistent with the total number of uninsured by income reported by the CPS survey (table 2).

Number of Children Per Adult

Table 7. Number of Children per Adult


Notes:

1. 12-17 Age Group - We try to account for teenage pregnancy in this row.

2. When age groups are not specific as in Wealthy People above, we use a weighted average to the 18-44 and 45-64 year old age groups.