Tyler Cowen has an interesting post on the comparison of the productivity of medical care in the U.S. vs. other countries. One brief excerpt:
Americans pay more but get better health care in return. We die sooner because we eat too much and exercise too little, among other facts.
Among, the many interesting links in his post, there is a paper by Richard D. Miller, Jr. and H. D. Frech, III which says
We found that increased pharmaceutical consumption is both economically and statistically significantly related to increased life expectancy at the ages of 40 and 60…In contrast, we found no economically or statistically significant effect of non-pharmaceutical health care consumption.
As Cowen points out, if you combine these findings with the fact that most drug research is done in America, then the United States deserves the credit for most of the productivity in health care not just here but around the world.
My cynical belief is that even if this were true, and could be demonstrated, many people would still favor drug price controls. People resent drug company profits, and they take drug research for granted. This would illustrate a phenomenon that I called Hating the Solution.
For Discussion. How much influence do you think econometric studies should have on people’s opinions about which is the best health care system? Would you change your mind if a statistical study contradicted your opinion?
READER COMMENTS
Eric Krieg
Apr 22 2004 at 8:48am
>>Would you change your mind if a statistical study contradicted your opinion?
My opinion on health care seems to change with every new econometric study released!
American longevity is a result of much more than just too many calories and too little excercise. We live in a stressful society. I’m changing jobs now, and will probably move as a result. It is incredibly stressful. I could totally become a drunk as a result. Neither is exactly good for my health!
So, to have some Euroweenie boast about his country’s health care system and longevity, while living in his ancestral village, working at his job-for-life, never worrying about getting laid off or having to find a new job to keep his skills fresh, it’s just a little too much.
dsquared
Apr 22 2004 at 12:19pm
Arnold: there’s a real missing link in your argument. “Pharmaceutical consumption” does not mean “Consumption of recently discovered pharmaceuticals”.
(and on a technical note, surely pharmaceutical and non-pharmaceutical care are strongly correlated. I’d be very suspicious about estimates of the significance of individual coefficients in a regression where the independent terms were so highly correlated).
Arnold Kling
Apr 23 2004 at 8:02am
D-squared: I am one of those people who would be reluctant to change his mind based on an econometric study. The paper on pharmaceuticals supposedly accounting for most of the health improvement is of the type that I deem suspect.
K Harris
Apr 23 2004 at 8:36am
The assumption in the question seems to be that econometrics answers every question, isn’t sensitive to the assumptions of the researcher or the structure of the model. So…no, I might not change my mind all that quickly in the face of yet another study.
I would also think that some of the objection to supplying drug company profits that fund research is outside the realm of the of the sort of statistical testing you suggest. If I pay the higher price, I am subsidizing users in drug-price-controlled economies, if not directly, then at least in their future access to better drugs. There is no incentive to change that situation if I don’t put up a fuss.
I also like the point that it is not necessarily new drugs which provide the bulk of medical benefits. The arsenal of pain-killers, antibiotics, blood-pressure reducers and so forth is already large. How “productive” is a drug few can afford, which is a substitute for a drug already on the shelf?
Eric Krieg
Apr 23 2004 at 10:04am
>>I also like the point that it is not necessarily new drugs which provide the bulk of medical benefits.
You could say the same thing about any class of goods. Do we really need new car designs? Why not stop all auto R&D? Why do we need new computer designs? Most people don’t use the computing power that they already have.
New desings are usually a marginal improvement over existing products. This doesn’t mean that new products don’t make our lives better.
luke
Apr 28 2004 at 10:31am
I’m just a lowly undergrad econ major. I’m still getting into the full mode of analyzing econometric studies. but I place high value statistical analysis, when it is conducted correctly.
I could change my opinions based on econometric studies if either I did the study, or the method of the study were explained in enough detail for me to believe that assumptions and biases were not present.
the latter could prove to be easier, since I would probably only take on study material that conforms to my own biases and assumptions.
Comments are closed.