Medicaid is the largest means-tested welfare program in the United States. Jointly funded by state and federal governments, its spending grows relentlessly whether the economy is adding or shedding jobs. Its ostensible purpose is to ensure access to medical care for households without enough income to pay for it. Yet new research suggests that only 20-40 cents of each Medicaid dollar improves recipients’ welfare. On the other hand, 90 cents of every dollar spent on the Earned Income Tax Credit (EITC) does so.

New research indicates why this gross disparity exists. In 2008 Oregon expanded Medicaid by lottery. The state randomly selected 30,000 of 75,000 on the Medicaid waiting list. This unique procedure approximated a randomized clinical trial, the standard in pharmaceutical research. This cannot usually take place for something like Medicaid expansion because government does not randomly determine who is enrolled.

Recognizing a unique research opportunity, a team of health-policy scholars formed the Oregon Health Study Group. Two years later it compared a number of outcomes for those who “won” the Medicaid lottery to those who did not. The group concluded that Medicaid increased the recipients’ use of health resources across the board and reduced the risk of large out-of-pocket medical expenditures. However, while it reduced depression, Medicaid enrollment did not improve mortality or any physical health measure.

The team has just published new answers to an even tougher question: How much does Medicaid increase recipients’ actual welfare? In other words: Does a $100 of Medicaid spending increase the dependent’s well-being by $100? More? Less? When we buy something, we make pretty darned sure it is worth at least what we paid for it.

Using a number of sophisticated economic techniques, the study group estimated how much Medicaid recipients themselves valued the program. According to the model showing the highest value, the average recipient would give up Medicaid for $1,576. Further, only $465 of the benefit of Medicaid is health care, while $1,111 of the benefit derives from freeing up resources for other consumption. That recipients’ behavior indicates they only valued their benefits at one-fifth to two-fifths of the money spent is a serious indictment of the program.

Observing that government spending per Medicaid recipient was $3,596, the study group estimated that the program reduced Medicaid recipients’ out-of-pocket spending by $489 and increased medical spending by $885, for a net cost to taxpayers of $1,374 per recipient. The difference, $2,222, represents the amount of government spending that was simply transferred to providers. Both $3,596 and $2,222 are significantly greater than the value of Medicaid as perceived by recipients ($1,576).

Yet Medicaid spending keeps growing, and is much larger than other welfare programs. I believe the root cause of the problem lies in a fundamental difference between Medicaid and these other programs. Citing data from 2011, the study group notes that Medicaid is the largest means-tested welfare program in the country, amounting to $425 billion versus $80 billion for food stamps, $50 billion each for the Earned Income Tax Credit (EITC) and Supplemental Security Income (SSI), and $33 billion for cash welfare (Temporary Assistance to Needy Families, TANF). Only in Medicaid do taxpayers’ dollars go directly to providers, rather than recipients as cash transfers or vouchers.

Medicaid spending is driven by providers, especially hospitals, which have relentless lobbying operations. This does not happen for other welfare programs. The National Grocers Association does not lobby to expand food stamps.

On the other hand, grocery stores are not run by religiously affiliated, nonprofit societies, as are most hospitals. The study group found that 60 percent of Medicaid spending comprises transfers to such providers for care they already give, rather than new care. These hospitals pledge to take care of indigent patients, but then lobby state and federal governments to underwrite those costs. While this plugs a hole in hospitals’ income statements, it is still uncompensated charity care from society’s perspective. It has merely been shifted from local communities to taxpayers nationwide. There is no reason to assume this is socially beneficial.

Medicaid spending could be significantly reduced, with no harm to recipients’ health, if the spending were converted to vouchers or something similar, instead of being paid to providers directly.