A new study on the Health Affairs one of the authors is with the Center on Budget and Policy Priorities.

I first looked at the paper willing to take them at their word. Lower costs are not necessarily a good thing, whether those costs are administrative or care-related. Lower administration costs mean no monitoring of who gets paid for what, like community support services. Lower care costs can limit access by driving providers away.

The authors, however, never get to the question of why Medicaid has lower costs. Their method of getting to lower costs is convoluted enough.

They recognize that low-income adults on Medicaid have higher medical spending than those with private insurance or no insurance. Those on Medicaid also are more likely to have “fair or poor” health, chronic disease, or had a child in the past year. Less than half of those on Medicaid are employed, compared to two-thirds of the uninsured and three-quarters of those with private insurance. Instead of examining the selection bias that these differences might reveal, the authors use “multivariate analysis and simulation methods” to control for the very health and demographic characteristics that might lead people to sign up for Medicaid instead of private insurance or to go without insurance.

Once they get rid of these important differences, they conclude that Medicaid results in lower spending. This is like comparing the John Locke Foundation research staff with the Boston Celtics, controlling for height, age, physical condition, and hours of practice then saying that the Locke Foundation would be a better basketball team. You control out the important factors and end up with meaningless data. I’d still put our big man D-Bakst against KG any day.

John Goodman of the National Center for Policy Analysis makes clear that Medicaid is already expensive, and getting more expensive, for states.