4 ms·
I really like this model and framework - most people underestimate the impact of transaction costs, and I think it's partially why many companies overestimate t
by dworin 13y ago
I really like this model and framework - most people underestimate the impact of transaction costs, and I think it's partially why many companies overestimate the value they'll get from outsourcing.
But in the real world, I've worked with dozens of leaders and professionals, and I don't think I've ever seen someone with an over-delegation problem, where they had handed off too many tasks that they should have been doing themselves. Most have the opposite problem: mistakenly believing that delegation will take longer, or will result in lower quality work.
Delegation doesn't just help the person delegating, it helps the person receiving the work as well. It gives other people the opportunity to handle new challenges, feel ownership over their work, learn new skills, and apply what they're good at.
- socalnate1 13y ago"But in the real world, I've worked with dozens of leaders and professionals, and I don't think I've ever seen someone with an over-delegation problem" Couldn't agree more. I actually think our brains tend to over inflate the transaction costs, not ignore them completely.
- johndcook 13y agoI've seen leaders who don't delegate but abdicate. They hand over a project completely, refusing to make decisions that only they are authorized to make. It's as if they're saying "Guess! And if it turns out poorly, I can say I never told you to do whichever course you take."
- jonnathanson 13y ago"most people underestimate the impact of transaction costs" I think most companies underestimate the impact of transaction costs, but most managers overestimate them. Both problems result from a failure to see the issue played out over the long term. In both cases, people or entities only consider the immediate to near-term impact of the delegation. A company will look at the present value of outsourcing, which is often highly positive, but underestimate the opportunity costs over a longer horizon (lost institutional knowledge, higher marginal transaction costs, commodification, etc.). A manager/individual will be hesitant to delegate because he or she only looks at the immediate hassle of having to train someone and get them up to speed. "It'll take Bob six months to really understand this track of work. It'll just be quicker if I get this done." In this case, the manager fails to understand that delegation has high short-term costs but low long-term costs. The learning "curve" is more of a step function. Once Bob is fully competent at Task X, delegating Task X to Bob will create efficiencies for his manager. It's starting Bob off on the task that's the hard part. His manager basically has to invest efficiency: "spending" some efficiency now in order to realize a net gain in efficiency later on.