Analysis: Politicians Using Flawed Data on Uninsured Population

Published December 1, 2004

Prior to the November election, presidential candidate Senator John Kerry (D-MA) wrote in an October essay for Health Insurance Underwriter magazine, “Roughly 45 million American have no health insurance at all.”

Ever since the U.S. Census Bureau released its August 26, 2004 report on the nation’s uninsured population, politicians have used the data to make health insurance policy decisions, and single-payer activists have used the data to lobby for government-mandated or -administered national health insurance. By the Bureau’s own admission, however, the data are incorrect: The number of uninsured is greatly overstated.

After a closer look at the data released in August and a consideration of analyses performed since then, a more accurate picture of the uninsured problem emerges. The good news is that the number of Americans without health insurance policies is not indicative of a crisis requiring more taxpayer-funded government regulations and mandates. The bad news is that policymakers are relying on extremely flawed data.

A Million More Insured

According to the Census Bureau’s methodology and published numbers, there were 44.9 million Americans without health insurance for at least a part of 2003, an increase of approximately 3 million over the previous year’s figure. At the same time, however, the total number of people with insurance increased by one million, to 243.3 million.

Devon Herrick, senior fellow at the National Center for Policy Analysis (NCPA), noted, “While it is true that the number of uninsured has grown, it is equally true that the number of people with insurance has grown steadily for the last 15 years. Despite recent economic hard times, there has never been this many people with health insurance.”

Although there were three million more people uninsured in 2003 than in 1996, the percentage of the U.S. population that is uninsured remained the same, at 15.6 percent.

Possible 15 Million Overcount

On numerous occasions, Census Bureau officials have acknowledged the uninsured number is inflated because the Bureau reports as “uninsured” those adults and children who are eligible for Medicaid and the State Children’s Health Insurance Program (SCHIP) but are not enrolled.

As Herrick observed, “Being uninsured in America is often a matter of choice. Most uninsured people either can afford health insurance or qualify for government-sponsored health care programs; they just choose not to enroll.”

While the Centers for Medicare and Medicaid Services (CMS) reported Medicaid enrollment at 51 million in 2002, the Census tabulated only 33 million, a difference of 18 million people. This same kind of undercount happened again in 2003: The CMS reported 2 million people became eligible for Medicaid, but the Census Bureau recorded only a 350,000 increase in Medicaid enrollment.

This is no minor statistical snag, as the Census Bureau reports there are more than 15 million “uninsured” individuals in households with less than $25,000 of income. Many of these individuals meet the income test for Medicaid or SCHIP eligibility, but they are not technically enrolled and are therefore considered by the Census Bureau to be uninsured.

However, as soon as a person who is eligible for Medicaid, but not enrolled, enters the health care system through a hospital or clinic, he or she is automatically enrolled into the Medicaid plan. Therefore, counting this population as “uninsured” distorts the data significantly, since these individuals can enroll at any time and have their medical expenses paid whenever they require health care.

The social policy implications of this overcount are important because the inflated numbers send the wrong message to politicians. One of Kerry’s health care proposals singled out the Medicaid population for special legislation. His campaign literature stated there are “millions of uninsured children who are eligible for health care coverage under Medicaid or SCHIP but are not enrolled.”

Given that no one who is eligible for Medicaid can be correctly described as uinsured, spending millions of dollars to enroll these people would do no one any good at all. These folks are not uninsured.

Many Unaware They’re Insured

In addition, many people may not be aware that they or their children are covered by a health insurance program, either private or government, if they have not used covered services recently. Thus they fail to properly report their insurance coverage.

Research I conducted in 1999 while at the Center for Advanced Social Research at the University of Missouri-Columbia discovered many Medicaid recipients say they don’t have insurance coverage, but when interviewed by a trained researcher and asked if the government paid for their medical care, they respond in the affirmative. When asked if they remember the name of the program, they cite Medicaid and sometimes Medicare. Many interviewees said they don’t consider government-run health care to be insurance because they do not pay premiums and often have no co-pays or deductibles.

Also inflating the uninsured figure is a sizeable increase in the nation’s immigrant population. Roughly 9 million documented and undocumented aliens are generally included in the Census estimates. Many immigrants hesitate to participate in a government program of any kind, for fear of establishing a paper trail for immigration and national security authorities. Cultural mores, folkways, and language barriers also conspire to keep these people uninsured.

Middle-Income Uninsured

Interestingly, the Census data for 2003 show almost 15 million uninsured people in households with annual incomes above $50,000, with 7.6 million of them in households with incomes of more than $75,000. That is certainly adequate income to afford health insurance in most states.

Another 18 million of the uninsured are between the ages of 18 and 34, and many people in that age group voluntarily take a pass on buying health insurance. According to Greg Scandlen, director of the Center for Consumer Driven Health Care at the Galen Institute, “This attitude is actually encouraged by ‘guaranteed issue’ laws in many states, which assure individuals that they can buy health insurance after they become ill or injured.”

Dueling Surveys

Data from three federally sponsored national surveys–the Survey of Income and Program Participation (SIPP), the Medical Expenditure Panel Survey (MEPS), and the National Health Interview Survey (NHIS)–also seek to make an accurate count of the nation’s uninsured population.

All three surveys conclude that at any given time during a year, being uninsured is a much smaller problem than we are led to believe by the Census data alone. For example, only about 30 percent of the non-elderly population who become uninsured in a given year remain uninsured for more than 12 months. Nearly 50 percent regain health insurance within four months.

Writing in response to the new Census data, Dr. Kirk A. Johnson, senior policy analyst for the Center for Data Analysis at The Heritage Foundation, highlighted the problem of taking the statistics at face value. The Bureau’s numbers may make for eye-popping headlines, he noted, but, “When it comes to health insurance, the Census Bureau’s own statisticians argue that SIPP [Survey of Income and Program Participation] provides a better measure of health insurance coverage than CPS [Current Population Survey]. In a recent research report on the differences between CPS and SIPP in this regard, Census Bureau statistician Shailesh Bhandari wrote, ‘Since the SIPP collects monthly information and allows us to see changes from month to month, SIPP may be closer to the truth.'”

“In short,” wrote Johnson, “the CPS data provide an incomplete picture on poverty and health insurance in America. Policymakers would be well advised to look to other data, such as SIPP, to gauge what actually happens to people who fall into poverty or lose their health insurance. Only then will public policy be fully informed, and America can truly have an intelligent debate on how to better address these problems.”

The Census Bureau’s own admission that the CPS “is not designed primarily to collect health insurance data” speaks directly to the issue that the methodology used to collect this important information is not up to the task.

New Thinking Needed

Grace-Marie Turner, president of the Galen Institute, suggests that since the uninsured rate has been static for the past 11 years, policy analysts and policymakers need to get their thinking “out of the box.”

“We would be gloomy were it not for last year’s public policy changes that can begin to shift the balance back to private health coverage,” said Turner. “Health Savings Accounts will help. Broader ownership of health insurance is the ultimate solution.

“The percentage of Americans without health insurance has been almost level for more than a decade. In 1993, the year of the great health care crisis, it was 15.3 percent. Last year, a nearly identical 15.6 percent of Americans were uninsured.

“At what point do we start to rethink the problem?” Turner asked. “The number of uninsured is slowly going to get worse without policy changes.

“Adding 45 million people to government program rolls just isn’t an option,” she noted. “Refundable tax credits for the uninsured, deductibility of individually purchased health insurance, and new purchasing options are crucial to begin to give more people more options to buy affordable health insurance–insurance they can take with them even if they lose their jobs.”

Derek Hunter, a research assistant for the Center for Health Policy Studies at The Heritage Foundation, summed up the situation in his August 26 report, “Counting the Uninsured: Why Congress Should Look Beyond the Census Figures.”

“At the very least, the undercounting of Medicaid recipients and the undercounting of insurance coverage, as admitted and described by Census officials, demonstrate that the Census Bureau’s figures on the uninsured do not accurately reflect reality and may lead policymakers and the public to incorrect impressions about the uninsured.

“The issue of uninsurance is simply too important for its public face to come from an indifferent and inaccurate survey,” wrote Hunter.

Conrad F. Meier ([email protected]) is senior fellow in health policy at The Heartland Institute.

For more information …

see the Heritage Foundation Webmemos by Kirk Johnson, “The Data on Poverty and Health Insurance You’re Not Reading,” and Derek Hunter, “Counting the Uninsured: Why Congress Should Look Beyond the Census Figures,” available online at and