Tuesday, July 07, 2009

Peter Principle Simulated

Some Italian professors have tried to simulate the famous Peter Principle:

"All new members in a hierarchical organization climb the hierarchy until they reach their level of maximum incompetence."

http://www.technologyreview.com/blog/arxiv/23800/

The Peter effect arises from the practice of promoting the best performer at Level N to a position at level N+1.

The Italians also simulated two other promotion policies:

1. Alternately promote first the most competent and then the least competent individuals.

2. Promote individuals at random.

According to their simulations, each of these methods improves the efficiency of an organization over the Peter method.

My thought: They could try promoting on the basis of who is most suited for the next level job. Duh!

Or maybe they could try not mixing the concept of "promotion" with that of "reward."

Or maybe even getting rid of the hierarchical notion altogether.

The Paul Principle

As a relevant post-script for my audience, they might want to look into the "Paul Principle," proposed by Paul Armer, who, like me, started out in computing as a desk calculator operator (or "computer" as we were known back then).

"People become progressively less competent for jobs they once were well equipped to handle."

Paul proposed his law in 1970, the year after Peters proposed his. Paul claimed his principle was more relevant in high-tech fields, when the complexity of jobs grows faster than the people doing them. The Paul Principle has been virtually forgotten, but I think it is still worth some careful thinking by IT managers and consultants.

The Other Paul Principle

It seems there's another "Paul Principle," after St. Paul's treatise in Corinthians:

"Continue to provide people with what they need to succeed."

I suspect this management principle would also prove more effective at growing an organization than the Peter Principle.

Perhaps those old Pauls knew something that's still worth studying today.

4 comments:

Unknown said...

Partly tongue in cheek and completely seriously -

"The Mary Principle"

Never underestimate the potential or impact of ordinary people on a mission that they believe in.

-- Tom Hawes

Bruce Johnson said...

The Peter principle and the original Paul principle are related of course. My own version of the Paul principal is to note that the half-life of any particular technical skill is 18 months -- it may be shorter these days.
So a good number of IT folks move into management as a way of avoiding the challenges of keeping their developer skills up to date. Much unhappiness ensues. The best carry forward some key principles of design, project leadership and testing. Many however are left in a dreadful situation. No longer able to understand in detail their developers' work and sometimes fearful of others discovering that fact, they become solely focused on the things that are easiest to measure and most rewarded by their management: meeting schedules and managing budgets and damn the consequences.
Given the Paul principle, the challenge is to define a set of career paths for maturing developers that can really allow them to build on their experience.
The best example of that I've seen was a career development framework in Price Waterhouse that allowed peers as well as managers to assess consultants on a regularly scheduled (and pretty intense and expensive) process. People were assessed in 6 dimensions, (e.g. IT skills, Project Management skills, Business Process, and Industry knowledge) and assigned a rating from 0-6 in each dimension. In that framework, people could define their own path -- deeply focusing on one dimension or developing breadth in multiple dimensions. The results were pretty good: it was possible to identify those people with a breadth of experience and skill who might fill a role in a project that would otherwise require multiple people; and also possible to identify those with deep expertise in a particular dimension.

Brian said...

Jerry wrote:

My thought: They could try promoting on the basis of who is most suited for the next level job. Duh!

Or maybe they could try not mixing the concept of "promotion" with that of "reward."

Or maybe even getting rid of the hierarchical notion altogether.

(end of Jerry's write)

I could write a book on this! And yes, Jerry, I'll put it in my books-to-write queue.

From the top: few if any organizations promote according to explicit criteria. It's the old-school-chum/my cousin/etc. stuff. They have no idea who is best suited, and are afraid that someone will find out that they aren't, either. So it's highly subjective and unpredictable.

Hierarchical organizations. They have their functions, as in armies. Most Middle East armies, while emulating British uniforms and drill, are mostly tribal. So alongside of a clear chain of command, they have social networks, which usually impedes military discipline. The wars of 1948, 1956 and 1967 bear out that this doesn't win wars.

OTOH, the more egalitarian US and European armies (not forgetting the Japanese, who modeled on the British), with definite chains of command (hierarchies), have been more effective in war, which is their purpose.

On the subject of military promotions, the US model is "automatic up to grade O-4 (Major or Lt. Commander)". One may note that Captain Queeg was really a Lt. Commander, and only "captain" by his command of a ship. This is where many military careers end or stagnate. After that, it's about "leadership" which as Jerry can tell you, isn't really clear cut. But it's more definite than in many civilian organizations.

(In one my several novels-in-progress, a Major has just been passed over for promotion for the third time, which is the end of the line.)

Recently, I've been having a conversation about "teams". My take is that some heirarch must tell the guys/gals that they are the team and they may organize themselves as they see fit, subject only to the teams' being responsible for results.

Unknown said...

Did they run a simulation on:

. Always promote the least competent person?

This is an old principle with the acronym FUMU (Foul Up, Move Up).

The concept is to promote your problem people so they will go away. It works in the short term as the problem people do leave your office. Of course they move on to ruin someone else's office.

The concept of promoting people who will do best at the next level is too hard to implement ;-)