the primary education system is in trouble. We make everyone spend a significant portion of their lives sitting around and unhappily learning stuff they won't use and that they won't appreciate in retrospect.
Unfortunately, it seems pretty sensible given its design constraints. This is not to say that someone creative can't think of a better system under those constrains, but it does make it a difficult problem.
Goals (in descending order of importance):
Design constraints:
First let's recognize that these constraints were not the ones under which humans evolved, so we can expect difficulty. I don't know much about it, but i've been told that the first constraints are unnatural in that in many tribal societies the kids can mingle with working adults somewhat, particularly in their preteen years.
I also note the absence of a constraint that is quite important in some other areas of life:
Now here's some difficulties associated with the goals and design criteria:
Goal "Teach the student "how to think" and "how to learn"" is best done with personal instruction with no very specific set of pre-planned activities (but possibly a library of potential activities), although maybe it can be done (i'm not sure), very inefficently, with a large "class" of students undertaking prespecified tasks.
I don't know if there is a way to measure the results of "Teach the student "how to think" and "how to learn"". There probably is a way, but i don't know what it is, and it may not even have been discovered yet.
It's hard to learn these sorts of things if one isn't genuinely interested, although not impossible.
The student must spend a lot of time having common experiences with and interacting with a number of studentren who are about the same age as they.
Many students find this boring, which isn't surprising, since this is "unnatural" knowledge that we didn't evolve to want to learn.
Some of these things will require effort to learn, so the student must be motivated. (note: one can have motivation without inherent interest, e.g. if something is boring but makes you rich).
These are various specific things to be learned here, e.g. reading, writing, arithmetic.
This kind of stuff is hard to teach by rote, and is best taught via human interaction rather than by impersonal communication channels.
This kind of stuff is hard to grok if you have no experienced it at least second-hand.
Many students find this boring (which is a little surprising in the history case, actually, since tribes have stories; but it's not surprising in the other subjects for the same reason given above, namely, that they are "unnatural").
Since you are doing this for society's benefit, not for the student's, it's not surprising that it may be hard to motivate the student if they don't have inherent interest.
It's going to be harder for strangers to do personal instruction or to motivate kids.
The kids will meet at least their teachers, so they'll get to know some adults from outside their family.
This means you have to concoct an artificial, unnatural cocoon, which is boring and hence unpleasant, and which also makes it difficult to provide hard evidence to students that whatever you are teaching is actually worth learning.
Someone has to babysit the kids.
...
The government doesn't have much money for this, and it won't get much, either, since many taxpayers don't have kids and resent being taxed to pay for others' kids, but there are also many parents who feel that parents should not be taxed more than non-parents -- the compromise therefore involves taxing everyone equally, but much less than the parents would like.
When one is running a mass organization one must be able to quantify results so that one can hire/fire/promote/manage/strategize by numbers -- no other reliable way to avoid non-performance in large organizations has yet been found (this is received wisdom rather than something i can support with evidence btw --- btw the explanation usually given is that for sociological reasons large organizations tend to be consumed by internal politics/bureaucracy unless constantly, harshly controlled by an objective stimulus signal). Therefore one must be able to test the learning of students, and optimize the learning process (including hiring/firing/promoting educators) to maximize test scores.
you can see where this is going, though. Many of the implications of the goals contradict the implications of the design criteria. One that is oft-talked about today is "teaching to the test". This is required by the design constraint of being run by a large organization, but unfortunately this means not being able to effectively teach certain subjects, such as "how to think/how to learn", which is the primary goal. Note that by my reasoning one should either accept something like school vouchers, or one should give up with trying to teach anything which cannot be effectively "taught to the test".