The theory behind outsourcing is that it enables specialization: you do one thing really well and let others do the rest.

Some of the more frustrating customer service experiences happen with entities where there's limited outsourcing and specialization. As I've written before, airlines do way too much. They market their brand and flight routes, they handle reservations and bookings, they maintain aircraft, they deal with luggage. More airlines should do as they do with their regional jet business: focus on something and outsource the rest. In the regional jet example, the big airlines handle reservations and ticketing and outsource the actual flying of planes.

City governments are another example. They try to manage parks, sewers, potholes, utilites, and more. Yes, a government entity, as the sole provider of police, fire, roads, and a few other things, will always be more diverse in scope than any sane for-profit corporation. But many governments still do too much beyond the core essentials, and are not able to do any one of these things very well.

One way to think about improving complex, ill-performing products, services, or experiences is to see whether there's a way to unbundle it and allow greater specialization. Arnold Kling applies this approach to improving higher education. Specifically, he thinks schools should separate the task of evaluating students' work from the task of teaching the concepts. Here's the background:

In the legacy education model, teachers combine coaching, feedback, and content delivery. By coaching I mean advice, guidance, and encouragement. Feedback includes formal grading as well as informal praise and criticism. Content delivery includes lectures and reading assignments.

Perhaps the key to radically changing education is to break up those functions.

1. The coach should be someone who knows the student well, who can relate to and motivate the student, who can recommend a good educational path, who takes account of the student's strengths and weaknesses, and who stays on top of how well the student is doing relative to the student's ability.

2. The formal feedback can come from strangers. Students can solve problems or write essays and have these graded by a separate service.

3. The content delivery should be "pulled" by the student rather than pushed by a teacher. For example, a student and a coach could agree that the student should learn statistics. The student then selects a statistics curriculum and works through it. The Khan Academy lectures on statistics are particularly good, in my opinion. But Carnegie-Mellon has a good on-line stats course, also. My guess is that, overall, there is enough content on line to obtain a world class education.

Then, Arnold writes:

A few months ago, Ben Casnocha wrote,

"Maybe 5-10% of high school high achievers should pursue higher education without attending a four year traditional college. This is the "Real Life University" option for entrepreneurial spirits. This is for folks who can learn a lot on their own, can assemble mentors and advisors to guide the process, and most of all find their creativity smothered by drudgery of school — or otherwise are on a trajectory higher than what college can offer — and therefore need an alternative path." 

His estimate of the percentage may be high, particularly in the near term. But that is the group that I wanted to aim at in my post on schools without classrooms.

Anyway, one important issue with alternative education models is interfacing with the legacy credential system. If you take a course from an alternative college, how can you get the credits to transfer to a traditional college or translate into a credible degree?

Arnold's proposed solution: A Means A.

A Means A solves the problem of credibility and comparability of grades in courses taught at different institutions of higher education. The innovation is to separate the grading process from other aspects of higher education. For any college-level course, A Means A will devise an appropriate exam and use independent professionals to grade the exam, according to transparent, standard criteria…

A Means A will extend the reliable, independent grading model of the AP exam to a broad spectrum of college-level courses. However, while the AP program compels instructors to "teach to the test," A Means A will "test to what you teach." That is, A Means A will take course objectives as given by instructors. It will design and grade tests that align with the objectives of the course.

It's a great thought. And it looks like one university is actually implementing part of it.

As a business opportunity, Arnold identifies the risks with A Means A, Inc. A company that promises to accomodate the idiosyncracies and variance of different schools' curriculua will have a hard time scaling the grading process in a cost-effective way. And making the credential have currency in marketplace in the early days will be tough. So while I am not so sure of the business opportunity, I think the high level prescription of unbundling is spot on. There are probably good business opportunties along these lines for education entrepreneurs–just need to brainstorm and iterate a bit more.

What Arnold has done with his A Means A post is bring to the table very specific ideas for improving the education system–not vague griping. And he aims his provocations directly at entrepreneurs–not policy wonks or politicans. A refreshing and useful approach.