At the most selective schools, a 2003 study found, just 3 percent of students came from the poorest socioeconomic quarter of families, while 74 percent came from the richest.
It is a fair guess that even among the 23% of students at selective colleges from the middle 50% of the socioeconomic spread, selective students are much more common among those near the top of that range than those at the bottom.
A mix of selective private and public colleges and universities (including the University of Michigan, of which I am a law school alumni), will not be recruiting for transfer students from community colleges to address the inequity.
A 2005 Department of Education study found more than one-third of 12th-graders in 1992 who went first to community college and earned more than 10 credits eventually transferred to a four-year college.
Bottom line: The "everyone can go to Harvard" myth, which selective colleges work very hard to propagate, is not terribly close to reality.
Yes, a select few less well off people make it into a selective college. But, it is the rare exception, for a variety of reasons ranging from poor preparation for college to an inability to afford an expensive institution. Given the way financial aid works (the least well off get the largest share of grants, those in the middle get work study jobs and large loans, and those who are most well off get token merit scholarships at best), I rather suspect that poor academic preparation is a bigger factor than financial inability. Recruiting from community colleges may make a small dent in the overall trend, but I suspect it will be small.
It is fair to note that one of the usual suspects, education funding inequalities in absolute dollars, probably isn't responsible for that poor academic preparation. For example, Denver Public Schools don't send many students to selective public and private colleges considering its size. But, this is not because per pupil funding is lower in DPS than in other districts. DPS is typical of the state by that measure. The districts with extraordinarily high per pupil funding turn out to be not the suburban school districts that routinely send many children to selective institutions, but small rural school districts that have small class sizes and high overheads, even though geographically the districts are far larger than their urban sister districts.
The problem is that it costs a lot more time, money and commitment from the system to get a kid whose first language isn't English and/or a kid who has grown up in poverty or near poverty to reach a certain level of academic proficiency, than it does to accomplish the same kind of academic performance in the case of a kid whose parents went to college and who has grown up in an upper middle class setting. Most obviously, standard academic English is the native language of upper middle class children, while it is effectively a foreign language for kids growing up in poor households just as much as it is for kids growing up in Spanish or Vietnamese speaking households. It is hard to learn a foreign language strictly in school.
As another example, upper middle class kids also grow up in more stable, safe environments, and have more opportunities outside the classroom, because parents want to give their children everything they can to help them in life and money can purchase opportunities. I've seen a child's SAT application and college materials tossed on the street in an eviction. An eviction or two, which are entirely the fault of parents and other people, can do grievious harm to a child's prospects, and uper middle class kids don't face those kinds of barriers.
Equal opportunity is a nice slogan and an ambitious goal. But, it is a predictable certainty that equality of opportunity, measured at times when it counts, like upon turning age eighteen, will never arise from equal commitment of resources to every child in the time leading up to that point. The paradox of education policy embodied in legislation with names like the "No Child Left Behind Act" is that few people are willing to acknowledge just how massive an effort is required to really achieve this goal. Unless resources are tilted grossly disproportionately towards students who start out behind "equal opportunity" or "leaving no child behind" will never be more than slogans. Policy makers like to believe that there is an easy way out of this problem -- that some new teaching method can make everything right for free. But, the magic bullet is very good a disappearing in the face of careful scrutiny and a need to mass produce it.