New Research Discovers Way To Avoid Organizational Incompentence: Promote Randomly

Alessandro Pluchino and colleagues from Universita di Catania used an agent-based simulation approach to try to figure out how incompetence happens in organizations, and how to avoid it. This replays the Peter Principle -- "All new members in a hierarchical organization climb the hierarchy until they reach their level of maximum incompetence." -- and suggests that there really is something wrong with always selecting the best at level N in an organization and promoting them to level N+1:

[via Technology Review: Blogs: arXiv blog: Why Incompetence Spreads through Big Organizations]

But is there a better way of choosing individuals for promotion? It turns out that there is, say Pluchino and co. Their model shows that two other strategies outperform the conventional method of promotion.

The first is to alternately promote first the most competent and then the least competent individuals. And the second is to promote individuals at random. Both of these methods improve, or at least do not diminish, the efficiency of an organization.

Interesting idea that would be fascinating to see in action. What would be a suitable prize for the first CEO to implement such a policy?

The promotion at random idea is so wonderfully oblique: it sounds like Borges "Lottery In Babylon", where each year the station of all people in some fantastic country was decided by lottery.

Comments

[Christopher Herot]

There was a study a few years ago (unfortunately I can't find the reference) that said an IQ test was a better filter than an interview for predicting the success of new hires. It wasn't that the IQ test was so good but that the usual hiring process was so prone to bias, especially hiring in one's own image, that almost anything else was better.

It looks like this research shows that the usual promotion process is worse than random. What does that say for the future of the corporate world?

[Stowe Boyd]

Chris - It says that people that are successful beleive that it's because they are so smart, and that luck plays only a small part in the way the world works. They can rationally choose people, based on these smarts, and then those people are more likely to succeed as managers. All hopelessly biased.

[david cushman]

absolutely love what this tells us about large hierarchies :-)

[Craig Roth]

"is something wrong with always selecting the best at level N in an organization and promoting them to levl N+1"?

Hmm, is it that tough to get around the problem by selecting the best suited for level N+1 (instead of N) and promoting them? Can't you can get an inkling of how someone will do at the next level by looking at the bits and pieces of that kind of work they have to do at their current level (like how a programmer did some architectural work before promoting them to architect)? I must be missing something.

But if you want to ignore all signals, not purposely give people bits of workload for the next level to test them, and avoid making educated guesses then I'd agree random is the way to go.

[Stowe Boyd]

Craig - That's the conventional wisdom, which this research debunks.

[F0ul]

Insane!

Seems what this research shows is that promoting people is the wrong way to keep an organisation healthy - which I agree with.

What usually happens is that the longer someone is in an organisation, the more friends and influances they have, and the more political the promotional process becomes.

If you only employed people for task N then you might get somewhere - with no preference given to current employees - but employing on a random basis is no better - but at least it keeps researchers in a job!

[Craig Roth]

"Craig - That's the conventional wisdom, which this research debunks." No, it's not true that my proposal is what this research debunks.

This research debunks the idea that "The people who perform best at one level of an organization tend to be promoted on the premise that they will also be competent at another level within the organization."

I proposed making an educated guess at "selecting the best suited for level N+1 ". I never said that person is the best at their current level - indeed they may not be. N /= N+1. I proposed assessing the times what N+1 type skills were excercised and providing opportunities to demonstrate N+1 type skills.

You may say that's conventional wisdom (I'm not conventional enough to judge), but it's not what is debunked by the study.

It's pretty obvious that selecting someone who excels at one set of skills to a position requiring a different set of skills may not succeed. We don't need more research on that. I'd like to see research on what organizations have successfully done to reward people who are the best at level N without throwing their skillset in the garbage. These people need some way to get a jump in pay to match their increased skills and promoting them to manager happens - with their approval - because its the only way to get more cash.

BTW - When promotion procedures correctly pick someone who is most qualified for N+1 but isn't the best at N, note how often the people that are the best at level N complain about it. "I'm way better at xxx then she is and she's going to write my review?"