The Trump Administration’s announcement that it will look favorably upon Medicaid waivers to require employment or “community engagement” has drawn much criticism, as well as a lawsuit, for its seeming incompatibility with the purposes of health coverage. But for those who are familiar with the long history of attempted social engineering within the Medicaid population, the new policy provokes a different question: Is health care, like cash assistance, something that is so desirable that withholding it will change people’s behavior?
In fact, experience shows the opposite: it is frustratingly hard to get people to use recommended care. The unwillingness of many people with asymptomatic chronic conditions to manage these conditions has been called one of the “hard problems in the social sciences.” The numbers on medication adherence are illustrative. CVS Caremark used its claims databases to quantify medication adherence for various chronic diseases on a state-by-state basis. Let’s look at the numbers from Kentucky, one of the states seeking to require work for Medicaid.
In Kentucky, on any given day, only 81 percent of insured individuals with hypertension have their prescribed blood pressure medication in their possession. Only 65 percent of hypertensive Kentuckians have those pills in their possession 80 percent of the time, and only 64 percent get the prescription filled a second time. The percentages are even lower for Kentuckians’ diabetes, high cholesterol, and depression meds. Incredibly, the nationwide medication possession ratio for asthma medications is less than half.
Let’s be frank—unless you are seeking relief from acute pain or incapacitating illness, receiving medical care is no picnic. Many of us are told by a doctor that we have some unseen, unfelt condition that will eventually hurt us if we don’t take a pill. These pills often have unpleasant side effects; for some pills, such as diuretics, the intended effect is not pleasant either. There are logistical and cost issues involved in filling prescriptions.
This phenomenon has spurred numerous demonstration projects to try to incentivize Medicaid enrollees and parents of CHIP enrollees to follow the prescribed regimens. But the results have been disappointing:
- Idaho’s Preventive Health Assistance program offered discounts on CHIP premiums to get children in for well-child visits. The compliance rate without incentives was 23 percent; the premium discount increased it to 49 percent.
- West Virginia’s Mountain Health Choices program offered upgrades to an “enhanced” health insurance plan in exchange for agreeing to a wellness plan. Only 10 percent of eligible adults took up the offer.
- Montana offered the Diabetes Prevention Program intervention to Medicaid enrollees at risk for diabetes. The baseline take-up rate was 47 percent, which was elevated to 60 percent at sites offering up to $320 in rewards.
- Hawaii—by most accounts the nation’s healthiest state—offered incentives of up to $320 for diabetic enrollees to adhere to self-management guidelines. Before the intervention, about 60 percent of patients were up-to-date on their retinal eye exam and flu vaccination; after the program there was no change.
- Wisconsin Medicaid mounted a Striving to Quit program that offered smokers up to $190 to complete a smoking cessation program. Sixty clinics referred over 600 patients to the program; only half enrolled. Of all those who enrolled, 75 percent completed the program and 67 percent appeared in person to collect the final $40.
What we can infer from the foregoing is that there are two basic groups of Medicaid enrollees. One we can call Group A, for Activated or Adherent, is people who are motivated to comply with their treatment regimen. Group N, for Nonadherent or Not Yet Committed to self-management, is people for whom free health care is not cheap enough. Within Group N we might say that there are two subcategories: in Subgroup N1 are a small number of people who might be motivated to comply by a side payment. People in Subgroup N2 are those you literally cannot pay to accept free health care.
The fallacy underlying the Medicaid work mandate is the assumption that enrollees are in Group A, when in fact it appears that around half of Americans with chronic conditions are in Group N. Taking away health care that this group of people doesn’t use often enough would provide the state with little or no leverage to coerce its preferred outcomes. These Medicaid enrollees seem likely to disregard requests for documentation of their work status and allow their coverage to lapse.
Given that a greater percentage of able-bodied Americans are in the workforce than regularly utilize their health benefits, the theory of action seems to be: if you don’t finish your dessert, you can’t have any spinach.
A further question arises with people in Group A —are they motivated to comply with their treatment regimens at any cost, or only when it is reasonably convenient to do so? One suspects that many people adhere to their medications because their doctor’s office and pharmacy have convenient locations and hours, and cost sharing is low. If hurdles are erected, will they remain activated? And what of people in Group N who undergo epiphanies and decide to pursue healthy lifestyles? In 2011 Wisconsin was paying such people to access care—now the same state, under the same governor, would lock them out of coverage.
Health policy fads come and go, but the hard problem remains the same: failure to address chronic conditions leads to preventable complications, leading to disability. Aspiring social engineers have a big enough challenge to keep them busy without taking on new ones.