[te_2939485943]

Two of these items, health care and education, have been inflated in America specifically by poor policy choice (some of which was perhaps enacted for good reasons, but had unfortunate effects). So they might be more controllable if there is a will to modify the system (even if, in the former case, it will be difficult and require stepping on many toes that currently benefit from the mess).

The less said about the mess that is American health care, the better. It is the one area where monopoly effects plague almost every part of the system. Whether it is pharmaceutical companies charging monopoly prices for new drugs (knowing that consumers don't directly chose, or directly pay for, what is prescribed)... or often monopolistic hospitals conjuring up obscenely high prices, billed often deliberately confusing and obtusely, designed for insurers to negotiate down... or insurers (a quasi monopoly, since it's what your employer choses, and your employer has limited choice) themselves coming up with confusing plans with a myriad of exceptions, where even a fully insured person can end up bankrupt after a major heath scare.

In practice, the "Bennett hypothesis" (the idea that increased generosity of financial aid leads to higher tuition) is the most likely explanation that I see for high education inflation. Perhaps a symptom of just how loose things were could be seen with the "ITT Technical Institute" type institution that spent far more time recruiting students via slick advertisements and taking their student loan money, versus working with businesses and developing a solid education program that created employable students. A working system would've never let ITT Technical Institute last for as long as it did. I think the expanded loans were made with good intentions, but they unfortunately were not clamped down stringently enough.

It would be interesting to see this chart repeated for other countries. Many of them probably don't have the issue with higher education and health care that the US does, but perhaps one could find other interesting issues. Some of which is not controllable... and perhaps some of which is.