With Jacques Steinberg's piece in last Sunday's Times (Plan B: Skip College), the "Is college really worth it?" meme seems to be in full flower, in part because it's an interesting issue and in part because the media suffers from a fatal weakness for novelty and counterintuition. But most of these discussions suffer from confusion about what question they're actually trying to answer. In roughly ascending order of importance, here's how various people are framing the issue:

Is college for everyone? This is a dumb question. Of course college isn't for everyone. Just last week, the Post profiled 17-year old high school senior Bryce Harper, who definitely shouldn't go to college. Instead, he should (and will) become a professional baseball player and earn millions of dollars. The number of good career paths that don't require a college degree is small and shrinking but not non-existent. Some people start families, others aren't smart or hard-working enough enough to complete college-level work. Defining the question in absolute terms does little other than identify the questioner as a sloppy thinker.

Does everyone in college need to be there? Again, of course not. There are 19 million people in college; obviously some of them shouldn't be.

Is going to college a bad decision for some students? Sure. Going to college incurs time and money costs, and produces benefits of various kinds. There's no upper bound on costs so logically they can exceed benefits. Borrowing tens of thousands of dollars for a substandard nursing degree, for example, is a bad idea. The average lifetime earnings differential for college graduates still exceeds the average cost of college by a substantial amount (the exact figure is subject to debate) but those are just averages.

(Johns Stossel thinks that because there are some students on the wrong side of both averages—costs too high, benefits too low—this proves that college in general is a "scam." His article does prove something: John Stossel is a hack. This we already knew.)

Are too many students going to college? This is a question actually worth asking. The Times article cites several credible academics, plus Charles Murray, answering "yes." Economist Richard Vedder, whose work I find thought-provoking if not always convincing, notes that 15 percent of mail carriers have bachelor's degrees. "Some of them could have bought a house for what they spent on their education," he says. But the optimal number of postal carriers with bachelor's degrees surely isn't zero. That's because of the specific nature of the college experience.

Matriculating at a university isn't like buying a car. Anyone with enough money can buy the nicest car available, at any time in their life, regardless of what cars they have or have not bought before. College, by contrast, is a process and an experience associated with a great number of prior and subsequent contingencies: You can only go to college if you successfully engage in various previous activities, and various subsequent options are only available to those who complete college.

Some students, moreover, are far more vulnerable than others to the policy choices likely to result from our collective understanding of these questions. Statistically speaking, my daughter will almost certainly go to college. First-generation students, by contrast, along with those who come from from low-income backgrounds and bad high schools, stand at the precipice of non-attendance. The way we think about college matters for them in profound ways.

Which is why the conventional approach to higher education has been, and should continue to be, expansive. It's a cliche, but it's true: College opens the door to opportunity. Not for everyone and not always, but very often and certainly often enough. Crucially, there's no way to know for sure ahead of time exactly who will benefit. Attempts to do so invariably discriminate against the marginalized students noted above. So we accept some inefficiency and additional societal expense, because the net result is positive and the people who benefit the most on the margins from expansiveness need it the most and deserve it the most. We're a wealthy nation and a surplus of enlightened mail carriers seems low on the list of problems to solve. How many of them, in retrospect, regret their degrees?

It would be possible, of course, to carry an expansive policy too far. But as David Leonhardt notes, income data show that the returns in the job market to a college degree relative to lesser credentials have steadily increased even as access to higher education has grown at the same time. And we should take seriously the collective wisdom of millions of college-educated parents who consider no option other than giving their own children a chance for higher education. This isn't just about status and social norms; it represents a rational and highly informed estimate of cost, contingency, opportunity and benefit, all pointing in one overwhelming direction. For the last century America has led the world in expanding access to successively higher forms of  education. Does anyone seriously believe this was, in retrospect, unwise?

Vedder et al. do make some good observations about college credentialing. There's a lot to be said for developing more creative, efficient, and flexible ways of certifying what people know and can do and matching those credentials up with the emerging labor market—as long as it doesn't have the effect of shutting students out of future opportunities to advance further down the postsecondary path.

But in the end, a lot of those questions really come down to whether or not the solution to various difficult higher-education problems should or should not serve the narrow interests of institutions and people who enjoy disproportionate wealth and power in society and have already benefited from access to college themselves.

If a lot of students enter college unprepared, which they do, we can shut them out of higher education as lost causes, or we can do the hard work of fixing public high schools and investing more resources in the community colleges and open-access public universities that do most of the heavy lifting in postsecondary education.

If many students drop out of college, which they do, we can can pretend that this represents fidelity to high academic standards or we can starting holding colleges accountable for graduating a reasonable number of students as compared to peer institutions with similar academic missions and admissions profiles, and do a much better job of giving at-risk students the academic support they need.

If the ever-growing cost of college pushes more students on to the wrong side of the cost/benefit equation, which it is, we can pretend that skyrocketing tuition is an immutable force of nature, or we can create a more transparent higher education market where colleges have strong incentives to restrain costs and ban colleges that plunge their students into unmanageable debt from federal aid programs.

College is extremely important and more people need it now than ever before. It's noteworthy that the people who argue otherwise are in nearly all cases great beneficiaries of college themselves.