• What’s the Medicaid crowd-out rate?

    At least a few people might be wondering what the Medicaid crowd-out rate is. A larger number might be wondering what that even means. Medicaid is said to “crowd-out” private coverage to the extent that Medicaid enrollees would otherwise be privately covered if they didn’t have Medicaid eligibility.

    This is a relevant issue as we consider the implications of various cuts to Medicaid. With less funding, Medicaid is likely to cover fewer people. How many who would be cut from Medicaid rolls would end up uninsured? How many would find private coverage? The crowd-out rate tells us.

    Those concerned about uninsurance should want to know about crowd-out. Those concerned about state and federal spending on health care should care too. Those who make a living on private and public payments to providers (hospitals, nursing homes, etc.) definitely care. Perhaps private insurers would be interested. (Since much of the Medicaid population is enrolled in private, managed care plans, it isn’t clear to me whether private insurers would favor a high or low crowd-out rate.) Of course, policy wonks always want to know everything!

    Beyond knowing the crowd-out rate, it is worth knowing that it differs a great deal which sub-population of Medicaid enrollees one is considering. Those who have or might have access to employer-sponsored health insurance have a far different — and higher — crowd-out rate than those who would otherwise be subject to the individual market. Also, since the economy matters in terms of who can get a job, the crowd-out rate for those who might be able to obtain employer-sponsored insurance can vary over time. (See also my prior post on the rate at which low-wage workers are offered insurance.)

    As much as I’d like to unpack all the crowd-out literature and estimates for you, I don’t have time. However, three recent papers, two of which are ungated, do a very nice job of surveying the landscape. They are:

    Though the technical details of these papers may be tough going for some, the survey of crowd-out estimates is not a heavy lift. Give them a look if you’re interested in this topic.


    • It’s shocking how sensitive these estimates are to specific methods and sub-populations. I did a mini lit review recently and came up with 50-60% crowding out during the Medicaid expansions from 1987-1992 and 1996-2002 based on:

      Gruber J, Simon K. Crowd-out 10 years later: have recent public insurance expansions crowded out private health insurance? Journal of health economics. 2008;27(2):201-17. Available at: http://www.ncbi.nlm.nih.gov/pubmed/18206255

      Cutler DM, Gruber J. Does public insurance crowd out private insurance? David m. Cutler and Jonathan Gruber. Quarterly Journal of Economics. 1996;(May). http://qje.oxfordjournals.org/content/111/2/391.short

      Cutler has a useful summary table of 12 prior studies that found crowding out rates between 0 – 70%.