Complexity and capability determine organisational structure

Thoughts on organizations complexity and structure

by Anonymous - 26.06.2011 - 8:28am

I have recently come to this blog/ webite, and obviously belatedly to this topic. I have known about Jaques and work for over 20 years, and actually briefly knew him and his prublisher wife. While I found the white paper accurate as far as it went, it still left open room for misunderstanding, and for some of the attacks on Jaques in terms of being anti-humanist/ human possibility in his thinking and advocacy, wrong about the enduring natrue of cognitive complexity status. Part of this stems from the very complexity (and sometimes complicatedness) of how he described things and the coherent need to consider interconnections of various kinds for a requisite organization to hold together and perform as wished for. There is no way to cover it all, but I think I have a few clarifying points to offer not covered or made explicit in the white paper.

First the ideas of individual cognitive complexity and organizational strata are a little like chicken and egg. Second, Jaques was very specific that he could only insure that his (ideal) requisite system could only hold true (with rare exception) in a corporate environment, in which the definitive flow of accountabilty could be established - literally and legally - by the fact of incorporation itself. While it is clear that in today's tumultuous economic situation and the acts that led to it we have largely abandoned people exercising their authority to hold others accountable in deference to the good of the shareholders, that is a fault of bad faith, poor ethics, manipulation etc. outside the domain of Jaques's premises.

While it is true that time is a critical variable in strata, it is an indicator of some factors not elucidated in the white paper. Again, the strata are connected to cognitive complexity (NOT IQ). Jaques became aware of and could demonstrate that the first five levels of cognitive complexity corresponded to five levels of symbolic logic... and it is in this domain that what distinguishes strata is to be found. While the white paper notes different typical accountabilies by stratum, it doesn't note the inherent shift in complexity and concommitant ambiguity that make the qualitative differences that can yield quantifiable results. Level I, e.g. deals with manipulating and making decisions about individual level work. The time span connotes the longest period typically encountered that an individual is not checked on, decisions and actions reviewed/noted. More to the point, if six individuals have assigned tasks with discretion appropriate to their work (be it engineering or clerical, or other) there is no fixed mechanism to make decisions like: fire somebody, have people work in combination on some tasks, redistribute resources for work being done... i.e. something that affects both individuals and the SET of individuals (2nd level symbolic logic). While Jaques was querrolous to the end about 'self managed teams', I think if he were less rigid even he would see that in some ways they can accomplish much of level II work, but not all. What this means is that a) the accountability for Level I output rests with Level II, and b) the nature of the key tasks for Level II add value (from the corporation's perspective) to the work done at Level I, even if all are smart, bright, committed workers.

Here's why time matters for strata in ways it doesn't for indivdual cognitive complexity. It takes more time to "get it right" and/ or to see results of decisions the higher you go. At the highest levels (5 and above) people are granted decision authority on behalf of the corporation for decisions that may be quite expensive, and may take longer to accomplish and "know the results" than the decision maker's remaining tenure in office. Not always, but often enough to characterize the distinction between levels, there is less certainty/ more ambibuity/ unknowns/ unpredictables when considering a decision at a higher stratum than a lower. Cognitive complexity seems to be among the few reliable predictors that correlate with thinking processes that deal with factoring not just unknown variables, but complex interactions and possibilites for variance within and among them. Eventually, this also includes conidering interactions among variables in similar or mixed sets that may represent entirely different domains of knowledge. While a bright Level I worker (even one w/cognitive compexity 5) may have some of the wherewithall to think cogently about this, the organization has him/her working on more linear and predictable things, definitely not how his/her work will be affected four years hence by changes in technology, AND state of the economy, AND the effects of globalization, AND pricing and marketing strategy, etc. If you want to see how some of that works out, study the strategies and decision making between Toyota and American auto manufacturers 30 years ago, between Ford and other American manufacturers in the previous eight years, etc. Clearly Ford's decision in 2006 to borrow $18B for restructuring not a snap or risk free judgement even before havinmg to deal with the unforseen (unforeseeable?) economic meltdown that began in 2007, and yet be able to adjust plans within the decision framework and no bankruptcy/bailoout.. -.

On a related note, I would speculate that the reason so many American companies are focused on short term gains is that so many of their leaders don't have the full cognitive repertoire to make sound decisions for positive results far into the future, and/or that those who support them may not have confidence that they have such capability, and/or their talent isn't matched by people in place in the organzation that represent the requisite cognitive complexity to optimize each levels conbtribution to what the Level V, or VII leader need to depend on for execution.

The real reason for higher pay in multiples by Stratum is both the increased complexity and AMBIGUITY inherent in making a decision that requires both more organizational resources and more time.

While there's tons more... I want to address something about organizational complexity and structure from a prescient book written in the 1960's by J. D. Thompson. He was a professor at Vanderbilt who, unfortunately, died young, and his book "Organizations in Action" was framed as merely a prolegomenon to a robust theory. To put this in perspective, he was writing about the same time Douglas McGregor was, but was too young to have attracted a following. At the time computers were all mainframe, and anyone outside a computer center entered data via teletype or keypunch cards.. no monitors, no feedback until your computer run either succeeded or failed, and nobody was using computers for "routine" office work. I say this in advance because history has demonstrated how accurate some of his postulates were as organizations downsized middle management the more computer intergrated they became.

His fundamental question was: Is there an ideal organizational structure/ form for businesses? Despite the ubiquity of standard bureaucracy (albeit amoong slightly varying structures), he decided that there were other forms. More he tried to figure out what different forms would look like, how they would operate. and which situations lent themselves to different structures as "optimum" for them. He came up with some complex and sophisticated stuff, but I and only to represent the grossest of his theory. Decades before anyone wrote of core competence, Thompson wrote about core technology, meaning what was basic to running the organization.

He came up with three ideal types (and of course combinations among them): Long Linked, Mediating, and Intensive. Long Link meant that completion of central value added activity a) took a long time, and b) required complex coordination from dsiparate functions and suppliers.. Think of automobile manufaturing or building nuclear plants as exemplars. Mediating meant facilitating or providing for the transfer of resources/ something, especially among entities that cannjot effect the transfers themselves w/o difficulty. Think of banks, the then existing land-line-only AT&T, and aspects of railroads and trucking as examples. Intensive meant the core work of the organization depended more on what was in selected individuals' talents/heads than on any optimal way to factor them into any chain of command or advance planning for their action. Think of universities and hospitals as examples, though with different degrees of intensity.

None of what he came up with meant that hierarchy had to be done away with, but its nature and how closely supervised, regulated, and coordinated its employees were had to be vastly different. Like the Stratum 5 decision described earlier, long linking requires tight control on coordination, accurate prediction of future likelihoods in different ways, budgeting over time, etc., and competent strata execution below that level OKing a new car for delivery to the public five years hence requires that, including such things as buying steel futures, discerning in what ways public tastes will shift that affect styling, performance, etc. The only form of managing;controlling this with greatest assurance, for all its defects, is bureaucracy. Mediating does not require such tight control and checking in across the hierarchy, though some is necessary. It needs something less bueaucratic, and in some ways more cross-linked activity at various functions and levels without resorting to the delay and inefficiencies of checking for authorization up and down "silos". Computer technology massively changed some of the businesses that mediated, e.g banks and phone companies as more and more of what they delivered - and needed to do to manage delivery well - could be represented as information not something that required moving and storing and putting together hard material. More critically short-time span speed was very important. Intensive technology needs traditional bureaucracy only to the extent that an organization's size requires coordination of other resources to support the core. Not only are academics best suited to decide what research to pursue, they are uniquely qualified to decide whom to do research with, even in only once. Profound work has been done not just by individual minds, but by - say - an economist, and a sociologist, and a poltical scientist - investigating something.... and then never collaborating with each other again. In a hospital one cannot schedule emergency surgery, or even assign to leading surgeons who the others in the operating room should be.

What I've taken from my readings and experiences with organizational forms and complexities is that there is no one best solution, and there are certainly other ideas out there not included here, but I think Jaques and Thompson represent two of the best theorists.

• reply