I'm preparing for the fall semester at the Harvard School of Public Health, and I've been digging through references about the impact of fee for service and capitation on utilization. I researched this for a Harvard Business Review article a few months ago, and found data on rates of cataract surgery in a group that was transitioning from fee for service to capitation. In that instance, the rate of cataract surgery dropped a jaw-dropping 51% in one year.
But there's some potential demagoguery in that number. It's likely that the ophthalmologists knew that the "index" year was their last chance to get paid "extra" for doing more surgery, and they might have advanced surgery from the next year into that index year, leaving a gap in demand during the intervention year. This would have also made the apparent baseline seem higher than it should have been. To my knowledge, a followup has not been reported.
I happened upon an article from Health Economics last spring that uses the Community Tracking Survey data (which matches what consumers report in terms of the medical services that they used with actual data on insurance coverage) and showed that surgeons being paid fee for service was associated with an eye-popping 78% increase in rate of surgery compared to capitation. The article is especially robust, because there was matching of actual insurance coverage and provider payment methods, and the analysis was restricted to patients who had no choice of insurance plan to avoid adverse selection in the fee for service group. Here's a full-text link for Harvard users. Whether primary care capitation lowered surgery rates depended on whether there was prior authorization in place.
My take - this is further evidence that moving away from fee for service provider payment is a necessary element to lowering the rate of health care inflation.