The central claim in Chris Conover’s March 28 Medicaid column (“N.C. faces a ‘Sophie’s Choice’ on Medicaid”) is blatantly false.
Conover goes to great lengths to praise the methodology of the Oregon Health Insurance Experiment, but he then makes the entirely false claim that the experiment “found that Medicaid had no impact on mortality risk.” Actually, the experiment was not able to answer the question of whether Medicaid has an impact on mortality risk. This is radically different from finding that it has no impact.
The experiment could not answer the question of whether Medicaid has an impact on mortality risk because the research design was not sensitive enough. For example, an experiment would be too insensitive to find any impact on mortality if it lasted only a single day – there just isn’t enough time for Medicaid to have an impact in a single day. You would not conclude that Medicaid has no impact on the basis of such an experiment.
Because the Oregon Experiment does not answer the question of whether Medicaid reduces mortality risk, we are forced to rely on the other studies that Conover describes in his column. As he points out, these other studies do imply that Medicaid reduces mortality.