04-22-2017, 07:29 PM
(04-22-2017, 07:01 PM)Belsnickel Wrote: The issue is that in any research scenario, you're going to find that the results of the statistical modeling will show findings that are not significant at the p<.05 interval, maybe a few at the p<.1. The results aren't likely to show definitive causation because of the rise in costs prior to the bill's implementation and other contributing factors to the change.
Here we delve into the field I'm working on a degree in, which is public policy assessment/evaluation. The difficulty in measuring outcomes to policy is extremely high for things like this. Saying a program has a positive or negative result is easy, proving it is another thing entirely. All because of external variables that can be difficult to control for. I only wish I was further along in my studies, or I could explain it a bit better!
Your explanation was fine. I'm not t too behind you. I stopped after a masters course in cost-benefit analysis; the kind that uses the NPV to determine whether or not we should do a public project.
I don't disagree that there are other things at play driving up health care costs. I'm not surprised that it's hard to observe, isolate, and quantify theses variables in a regression set.
But that isn't grounds to dismiss the idea that the rise in demand is a contributing factor to rising costs.
