The short answer is: nothing.
The appropriate way to evaluate a system is based on what it is designed to do, and the education system is not designed to do anything. Rather, throughout 200 years of public education, it has been given a sequence of demands, and has responded by adapting organically.
Among other things, the system has been required to:
- prepare young men for factories and the military, and young women for marriage and domestic service;
- serve as a mechanism for upward social mobility (Dewey);
- produce an educated electorate (Jefferson);
- generate 'well-rounded' individuals;
- serve as a minor league for professional sports;
- fix major social problems (Head Start);
- provide enough science and engineering majors to keep the economy working;
- graduate most students, each immediately ready to be successful in higher education, or prepared for a job;
with no recognition that some of these goals are in direct conflict.
For example, demanding increased performance necessarily means that more students will fail to clear the bar. Alternatively, demanding increased graduation rates necessarily leads to grade inflation and lowering the bar. There is no way around this, as it is precisely the problem of Type I and Type II errors in statistical testing.
The over-emphasis on sports in some districts, and the effect on grade inflation, has been well-recorded. What is rarely noted is the effect on the other students. If unprepared student athletes, or others who seem to put no effort into their studies, still progress to the higher grades, what is the immediate incentive to work hard?
If teachers are trying to fix every social problem in their classrooms, where is the time for learning? In my experience, the more talented students tend to get less attention, because 'they will learn anyway'. This is a recipe for mediocrity.
In short, before we try yet another major overhaul of the system, we should perhaps first decide the goals to be met, and also check that those goals are actually achievable.