Does Health Insurance Insure?

There's a 2008 report called "The impact of health insurance on health" that is vital to briefly review; in three conclusions it neatly and pathetically summarizes the current state of US health insurance:

Research on why health insurance fails us offers little guidance.

1. Many of the studies claiming to show a causal effect of health insurance on health "do not do so convincingly because the observed correlation between insurance and good health may be driven by other, unobservable factors."
2. Convincing evidence demonstrates that health insurance can improve health measures of some population subgroups, some of which, although not all, are the same subgroups that would be the likely targets of coverage expansion policies.
3. For policy purposes, one needs to know "whether the results of these studies generalize."

The researchers concluded: "Solid answers to the multitude of important questions about how specific health insurance policy options may affect health seem likely to be forthcoming only with investment of substantial resources in social experiments."

- H. Levy and colleagues, University of Michigan, Institute for Social Research.

The above quote is from Health Insurance Week June 1, 2008 Section: Expanded Reporting; p. 59

A colleague responded to this asking if this isn't backwards logic. That is, he asks "Isn't the causal relationship an issue of access to the health care in the first instance because you have insurance as opposed to those who have no insurance and therefore no access to healthcare??"

Another retorts:

Jeffrey:

When there is a positive feedback with a negative effect, does it really matter what is cause and what is effect? Lacking insurance makes it more likely that an illness will go untreated and or become more serious. Having an illness makes it less likely that one will be able to obtain

insurance.

It is a cop-out to ignore the effect by glibly blaming "other unobserved factors". But it is typical of lazy academics who are interested in health only as fodder for their next grant. Interruption of a feedback loop at any point will stop the effect. Isn't THAT what he should have said?

What do you think? Isn't it a cop-out to keep asking for more research "because good health may be driven by other observable factors" or “same subgroups that would be the likely targets of” improved coverage policies? Why not just emulate the best practices of current insurers (there must be a few that do a good job) and health plans that seem to care about patients; find out what combination of accessibility, cost and quality leaves patients the healthiest and happiest?