• It always comes back to the RAND HIE

    This is a TIE-U post associated with Jonathan Kolstad’s The Economics of Health Care and Policy (Penn’s HCMG 903-001, Spring 2012). For other posts in this series, see the course intro.

    Talk about health policy long enough and there’s a good chance that the RAND Health Insurance Experiment (HIE) will come up. So it goes for this series of TIE-U posts as well. This concluding post pertains to the HIE. (For background on the HIE, see this post. For much more, see all posts tagged “RAND HIE.”)

    In Patient Cost Sharing in Low Income Populations (pdf), Chandra, Gruber, and McKnight analyze changes in plan copayments for subsidized plans offered to low-income populations in Massachusetts from October 2006 to June 2008. As in the HIE, a key question is, to what extent do people cut back on health care when it costs more out of pocket? Another question, not the focus of the paper, is, what are the health consequences of the cut backs?

    On the one hand, smarter plan design could help reduce the fiscal pressures associated with insurance expansion. But on the other, it is also possible that low income recipients are unable to cut back on utilization wisely and, consequently, experience hospitalization “offsets” as a result of greater levels of patient cost sharing. In particular, there remains a concern among many that higher cost sharing on primary care will lead to less effective use of primary care, worse health, and, consequently, higher downstream costs at hospitals (the so called “offset effects”).

    A round-up of posts related to the consequences of cost sharing is found on the FAQ page. Among them is one that points you to a synthesis paper by Katherine Swartz, in which she concludes that increases in cost sharing can be particularly problematic for populations with low incomes and/or chronic illnesses. Such groups are more likely to forgo health-preserving care, not just reduce unnecessary care. This echos findings from the RAND HIE, which found health impacts for low income, chronically ill study participants.

    Chandra, Gruber, and McKnight found evidence consistent with the RAND HIE.

    Regardless of the chosen sample, the estimated elasticities are very similar to each other, with magnitudes ranging from −0.162 to −0.346. […]

    Our estimates of the price elasticity of demand in a low income population point to elasticities that are very similar to those obtained in the HIE—a pioneering experiment that was conducted over 30 years ago. That our elasticities are so similar, despite tremendous structural changes in the composition of what accounts for medical care, is remarkable.

    And what about health consequences or downstream utilization of other services? The study didn’t explicitly focus on that.

    In the present analysis, the prima facie case for offsets is weak as we see reductions in the overall use of hospital spending. But our analysis has not explored the possibility of heterogeneity in offsets, and in particular a situation where most members reduce hospital utilization by a small amount, though there are certain subpopulations where the use of ER and hospital services increased substantially.



    Comments closed
    • So it looks like Robin Hanson’s view of the Rand HIE still seems to stand.

      How many diseases are their where early treatment is significantly cheaper than watchful waiting and latter treatment? And let us also keep in mind that errors in early treatment can also lead to complications that could have been avoided. Early treatment can also lead to false positive that can lead to problems.