Time Span of Discretion is what determines the size of the role, and not some measure of “complexity”.
Harald Solaas, who wrote a comment to “Does Requisite Organization Really Work Over the Weekend?“, has written an article entitled “Why Is Requisite Organization (RO) Theory So Difficult to Understand?.” In it, he relates the following story about working with Elliot Jaques on a Requisite Organization project:
…Jaques absolutely rejected trying to establish the size of a role by looking at the problems its incumbent had to handle. I remember him once offering the example of a general manager’s role that had all the appearances of parallel processing (stratum IV), with accountability for the coordinated functioning of several departments, that had been measured as stratum III. I asked him, “You mean the incumbent was a serial processeor?” “Yes”, he replied.
This really surprised me.
I had theorized that time-span of discretion was a proxy for complexity, since decision points continue to branch out geometrically as time goes on. Paul M. Cashman and David Stroll make the same point in “Achieving sustainable complexity through information technology: theory and practice”, a 1986 ACM paper that used Jaques’s SST, apparently using data from an earlier project. They make a statement for “time-compression”, where executives reduced “the time-span to cope with the increased rate of change in their environment.”
Donald Brooks, a KPMG Canada partner and one-time Harding Consulting Group consultant, states something even more scandalous in “Today’s compensation systems: rewarding the wrong things,” a 1994 article in Canadian Manager:
Let me illustrate a source of confusion. A strategic planner may prepare a five-year plan; the target time to complete it is five months. The TSoD [time-span of discretion] is five months, not five years. Doing this is not as easy as it appears. I have discovered in many organizations claiming to have clear role accountabilities that the time dimensions of tasks are only implicitly understood. It takes some work to shift people to this concept. Once they do, they also begin to see how muddled many of their accountabilities really are.
Many studies confuse TSoD with planning horizons. My own experience is that TSoD accurately measures role complexity. For those who use it, TSoD performs better than job evaluation point factors and with greater mutual understanding and reliability.
Jaques made very clear that managers cannot delegate their planning activities. However, if we take the time-span of the planning activity and map it to a role, we have a very short time-span of discretion. You can create a five-year plan in much less than five years. Does this mean that the plan can be created by someone with a 3 month time horizon?
I had thought that certain IT projects, because of their inherent complexity, required longer time-span of discretion because they required a certain mode of thinking (parallel processing). We break up the project into small subprojects of interim deliverables that are run consecutively and not concurrently. Each subproject has its own project manager. My experience (and my gut) said that this didn’t work: you needed someone who was seeing the entire project.
According to Solaas, Jaques vehemently disagreed.
Saying that planning can be done by someone who cannot think out to see the entire plan is akin to saying that I can construct a write out a Stratum IX argument because the act of writing it only takes a couple of days. If I can’t think in that way, what does it matter how short the time span of the task is?
Perhaps I am misunderstanding time-span of discretion or Solaas’s and Brooks’s arguments. I can’t see how Brooks was saying anything other than what he was saying, though.
Image Credit: Advertisement illustration. Via Library of Congress collection.