Nick Cronin wrote a very interesting piece on how technology enhances labor mobility: http://techcrunch.com/2012/02/04/labor-efficiency-the-next-great-internet-dis...

I have a different perspective on labor mobility... or more precisely, wether I as an employer would desire fungible talent. I don't.

Putting aside the functional skills that a given person has, there are important social skills that take time to assimilate. Getting communications and interactions and understanding right takes a fair number of interactions. It's not just something that happens instantly (okay, sometimes you find someone who "speaks your language", but that's not the norm, it's a glorious, but rare, event.) It takes time to cultivate a shared language, a shared set of norms, a shared set of expectations, and only then, can someone work best for a given organization.

The requirements for shared language (my shorthand for the broader set of social and cultural understandings) is necessary for a worker to work out well within an organization. For example, there are some people who won't show their code until, in their minds, it's perfect. Those people may thrive in one organization but utterly fail in another. This is also true of people that code incrementally (lemme toss this out, get some feedback, and try again.) In both cases (and granted these are extremes), different organizations will do better or worse with each type of developer.

But, you say, we can settle those kinds of things in an interview.

No, no you can't. Most organizations use similar words for describing their expectations and norms. Lots and lots of organizations look for "Rock Stars" and "Ninjas" and are "Agile" and the like. Yet, those words mean different things. I've seen companies that have 45 minute "stand up" meetings where the director of engineering gives out daily assignments. I've seen organizations that advertise for "Rock Star Coders" and are frightened and institutionally reject folks that I consider merely good coders because they are so much better than the existing pool of developers that hiring them would cause more friction than the value of the additional good code. It's only in the exploration of the relationship that the parties can come to learn about each other and determine if they have a shared language.

Yes, you say, that's all good and fine for "knowledge workers" but what about the replaceable humans who sling hammers or assemble cars or take care of our lawns, etc.

The vast majority of work that humans do requires judgment and understanding priorities and balancing those priorities. The folks flipping burgers at In 'n' Out make lots of judgments. The folks at McDonalds try to remove as many judgments as they can and they will do best hiring people who do not thrive on making decisions. I'm betting there's not a lot of cross-trading between folks who are successful and McDonalds and In 'n' Out.

In fact, I think one of the things that cost Detroit its leadership in the 1970s (leadership that it has yet to regain) is the view that the assembly line workers were fungible, not smart, not valuable. The mentality led to management/labor friction, anger, waste heat, and ultimately lost leadership. If GM and Ford and Chrysler has viewed every member of the company and the supply chain as a valuable contributor to the creation and delivery of products that make buyers happy, we would live in a different world. But I digress.

Yes, sometimes you want to have external professionals rather than have them on your staff. Lawyers are one perfect example. Two of many favorite corporate lawyers are Jon Gavenman http://www.cooley.com/jgavenman and John Hession http://www.cooley.com/jhession. I've interacted with a lot of transactional guys in my life, yet these two stand out as working well for me. That doesn't mean they would work out well for everyone. They each speak the same language that I do, but that doesn't mean they speak the same language as everyone. Neither of these guys are interchangeable with any other lawyer.

But the bottom line for me is that institutional knowledge is gold. It's the core value of any company. That institutional knowledge is not simply knowing the problem space, but also the institutional knowledge of the values of the company, what the priorities of the company are, and how the company deals with adversity. The only way to foster that institutional gold is through long term relationships and the idea that any worker, any member of the team is fungible reduces the value of the business.