At some point there are probably diminishing returns to the amount of $$ spent per student in a school system. But you can imagine that most suburban parents would have a hard time reducing the $$ spent on their own kids for the benefit of the greater good.
http://www.businessinsider.com/we-now-know-more-about-how-ne...
I would argue that the reason suburban students are better are most likely because their parents are also better, whatever being a 'better' parent means in this case.
Poorer communities often have a culture so different than the average HN reader that most of us can't possibly understand it. (I learned this as a result of being a foster parent)
If you were trying to rank parents wouldn't "prioritizes their children's education" be a factor in determining which parent was better? Just because something is a cultural difference doesn't mean we can't say it is better or worse.
As nice and easy it would be to measure a person's worth as a parent by how well their kids do in school, or how much they make as an adult (like this article), I don't think that's reasonable at all.
Look at a state like Connecticut, which features very affluent areas and very poor areas within. If you like at the top performing schools, sure the best ones tend to be in affluent areas, and the worst ones are in cities, but that's not at all how much money gets spent per student. For instance, the town in the state with the absolutely lowest spending per student, Wolcott, is one of the top school districts in the state, and even has a school that won the national "Blue Ribbon" aware for excellence last year.
There are also numerous studies (which I'm way too lazy to go find) that indicate the affluence of the parents to be more important than the affluence of the school on a child's school success (this article is about adult income, though).
I ask because I think it can be hard to generalize from an area like Connecticut that is really quite high population density and quite wealthy to the US in general (in CT a student might be a reasonable distance to 3 or 4 large high schools. In Nebraska, they might be a reasonable distance to 1 tiny school).
Keep in mind that CT is a bit of a sinking ship, with companies and jobs leaving, and the population often following it. It's not some place that everyone is flocking to, but it has a long history of being at the forefront of things like education and mental health.
I've also heard, from numerous people, that having a few years experience as a teacher here in CT will make getting a job teaching in most other places completely trivial.
The other cycle is self-reinforcing too: Low housing cost -> working class population -> less tax revenue -> worse schools -> reduced home demand -> lower housing cost -> even less affluent population ... ... ...
There doesn't seem to be an obvious weakest link, but I'd start with untying school funding from LOCAL taxes and from "school performance" which is code-word for "richer students". Allocate funds state-wide or even nation-wide to those schools with the highest need.
EDIT: Alternatively, let me send my kid to whatever school I want, regardless of which side of some artificial school district line I live.
Instead, in the 'good exurban schools' case, the evolution usually goes:
1. Rural, periphery area attracts affluent folks seeking bigger houses on cheaper land. High-end residential subdivisions are developed.
2. Increased tax revenue allows for good schools, nice government amenities.
3. Good schools attract additional residents. Residential population swells, commercial and industrial population remains low.
4. Residential brings fewer tax dollars per infrastructure cost than commercial or industrial. At some point, municipal costs (roads, sewer, fire, schools) begin to approach revenue. Services are cut. Schools decline, making residents eligible to contemplate moving to another area in step #1 or step #3. Those who can't afford to move to a newer area stay; they are thus on average lower-income than those who left. At this point, there is a danger of falling into the second cycle.
The evolution I laid out can be most readily observed in places where urban areas are not hemmed-in geographically, and easy access to cheaper undeveloped land exists, so new exurban nodes can be readily created. Right now, Northern Virginia (DC suburbs), Atlanta, Raleigh, Dallas, Phoenix, Salt Lake City, Kansas City, and Minneapolis are experiencing this in their urban areas. California yet again a special case because exurban areas are usually much, much further away than elsewhere (Santa Clarita, Temecula, Palmdale -> LA; Santa Rosa, Fairfield, Tracy -> SF/SJ).