By John C. Goodman
Co-authored by Linda Gorman
Originally posted at Forbes, May, 2017
The Republican health plan will kill people,” says Bernie Sanders. “The Republican Party is the party of death,” says the headline at Political dig. “Repealing Obamacare will kill more than 43,000 people a year,” says an opinion piece in the Washington Post.
Other writers have shown similar lack of restraint:
- “If 24 million people lose insurance, we’ll see more than 24,000 extra deaths per year,” writes Julia Belluz at VOX, adding that “There’s no question that the fate of the Affordable Care Act is a matter of life and death.”
- “House Republicans are … consigning thousands of people with serious illnesses to death by making them uninsurable in an era of unaffordable treatment,” writes Michael Jeffries in the Boston Globe.
- “Nearly 45,000 annual deaths are associated with lack of health insurance,” writes Charles Blow in the New York Times, adding that the latest Republican health bill poses a “death threat … for many Americans.”
The Washington Post gave Sanders Four Pinocchios for his claim. We’d like to give all these writers two dozen Pinocchios. Every claim they are making is based on “fake research” that has been thoroughly discredited by independent scholars.
By the way, this is all a rehash of the scare tactics used to justify passing Obamacare in the first place. For example, “Dying for Coverage,” a report by Families USA, made the astounding claim that 6 people died every day in Florida because they are uninsured. Eight died every day in Texas, 9 in California, and 4 in New York.
How was Families USA able to tally up all that carnage with such pinpoint precision? As we explained in 2009, these claims are based on a chain of reports created over two decades. Each preserved the errors in earlier efforts, misinterpreted or mischaracterized the original findings, and based its projections on data that are now more than 40 years old.
The chain begins with a paper by Peter Franks et al. in the Journal of the American Medical Association in 1993. In the early 1970s, people were interviewed (only once) to determine their health insurance status. In 1987, they were surveyed to see who was alive and who was dead. The researchers made the implicit assumptions that someone who was uninsured in the early 1970s remained uninsured until 1987 and that someone with private insurance was continuously covered over the same period.
The authors made no attempt to determine the causes of death. Auto accident fatalities, gunshot victims, suicides – all these were implicitly assumed to be affected by whether people had health insurance. Armed with these heroic assumptions, the authors concluded that being uninsured increased the probability of death by 25%.
One problem is that being uninsured is like being unemployed. It happens to many people for brief periods of time. Most people who are uninsured regain insurance within one year, and people rarely remain uninsured for almost two decades. Another problem is that the authors made no attempt to determine whether access to medical care was in any way related to the causes of their deaths.
All in all, these are not results a careful researcher can have much confidence in.
The next link in the chain was provided by the Institute of Medicine’s (IOM) 2002 report Care Without Coverage: Too Little, Too Late. It calculated that 18,000 deaths a year in the U.S. are attributable to a lack of health insurance, using the 25 percent result from the Franks study without comment. The Urban Institute simply updated the IOM calculations in 2008. Families USA applied the Urban Institute updates and the IOM method in its imprecise estimates of deaths from lack of insurance in each state.
June O’Neill, former Director of the Congressional Budget Office and her husband Dave, both highly respected economists, approached the question of how health insurance affects health from another direction. They concluded that uninsured people with lower-incomes were only 3% more likely to die over a 14-year period than those with health insurance and being uninsured had no statistically significant effect on anyone else. Later studies support this finding. Card et al. find that “Any plausible effect of insurance on health status in the general population will likely be small” and perhaps nonexistent.
Given that lifesaving medical care has historically been available to people in the United States even if they cannot pay, the relatively small effect of health insurance on mortality is not surprising. Furthermore, most people in the US have the income, assets, or credit to pay cash for commonly needed medical procedures.
It is often claimed that the uninsured are at greater risk because they inappropriately delay seeing a doctor. However, this problem is unlikely to be satisfactorily addressed by the Obamacare solution of enrolling people in Medicaid, which routinely rations care by making people wait. Roughly half of doctors are not taking any new Medicaid patients, and a third are not seeing any Medicaid patients at all.
The Oregon Health Experiment found that enrolling the uninsured in Medicaid produced no significant improvement in common clinical health outcomes. It agrees with a growing academic literature suggesting that Medicaid patients do worse than the privately insured with respect to morbidity and mortality. In fact, forcing large numbers of people to join Medicaid may be worse than no coverage at all.