The assumption is that health care consumers make a trade-off between the price of the care and its true benefits. This usual policy logic dictates that we use elasticity of demand (the change in quantity demanded divided by the change in price) for a category of care to set co-payments. The data indicates a high demand elasticity meaning that the care is of low value. If a small change in price results in less use, than the value of the care must not be very high.
But what if that isn't how people make decisions? What if people have biases in judging in making decisions? A particular difficulty with medical procedures is that many people tend to judge them much worse before than after the procedure. Their fear of the procedure makes them subject a bias again the care that is confirmed by the co-pay. People suffer from a great deal of confusion and anxiety when making health care decisions.
What if people aren't very good had judging the medical risks of various procedures? Could people myopically discount adverse health outcomes?
While there everyone agrees that co-pays create lower utilization rates, the important question is, do co-pays end up limiting overconsumption of medical care or does it lead to underconsumption of medical care. Do patients reduce the amount of care they purchase based on inaccurate assumptions of risk thus leading to higher long term costs for consumers as well as higher social costs for society?
This is particularly likely to be true in the case of low-income patients. A $50 co-pay may reduce overconsumption by high income patients while creating underconsumption for low income patient.
There are alternatives to simple using a meat ax on health care. Differential pricing could be valuable in this process. One idea would be to increase the co-pay as the number of visits increase. Another would be to set the co-pay as a percentage of income. Co-pays may just be another excuse to cut services to those of us who are less well off.