Do you work in or for an organization?
Do you think your organization could work better? Be more effective?
Let's talk a little about organizations, communications, and why big organizations act differently than little ones. Hopefully from this you'll be able to see new ways to structure the organizations you are involved with so they'll work better.
First, what is one of the major reasons for having an organization?
It's to let people work together for a common goal.
And, what is one of the things people must do with each other if they are to work together?
They must communicate.
And how much time must a person in an organization spend communicating?
It depends on the job and the size of the organization.
And how much time does a person have in a day? Does that depend on the size of the organization?
Definitely not!
So, the amount of time we have in a day is fixed, thus, the amount we can spend on any combination of tasks is also fixed. Interacting with other people takes time, and the more people involved, the more time must be spent in interacting. This time required for interaction can become one of the major inefficiencies of large organizations. It is a diseconomy of scale.
Next, I am going to explain a model for meeting times. Now, the purpose of this model is not to explain what is happening in your office, but to allow you as a manager to make some accurate predictions about what kind of productivity changes you can expect as you add or subtract people from your office environment.
As is common with most models, it is much simpler than the real situation it is trying to predict, but it is this simplicity that makes it useful because it makes the results easy to calculate.
The assumptions I make in building this model are:
Given this simple model of how people in an organization acts, let's see what we can deduce.
The problem we face in group communications is very similar to the problems faced when sorting information--the time it takes to sort information depends, among other things, on the number of items to be sorted. Unless special sorting routines are used, the time to sort will increase by the square of the number of items to be sorted. Unless special efforts are made by an organization to reduce interactions, the time taken up by interactions will increase in a similar fashion--it will increase with the number of combinations of the people involved taken two at a time.
=================================================================
Number of contacts between people in an unstructured group where every member talks to every other member.
People in group: 1 2 3 4 5 6 7 8 9
Contacts / Time: 0 1 3 6 10 15 21 28 36
=================================================================
[[[ 5 - Networked communications structures ]]]
Imagine this as the minutes it takes for a group to decide where to go for lunch in a strange city, and leave to do it. One person can decide instantly; two will take a minute; and nine will take 36 minutes to discuss the situation and follow up on the decision.
Next, a demonstration of just how important this can be. Suppose you have two groups of people working for you. In A Group the members meet with each other for a half hour each day. In B Group they meet for an hour. Both groups currently contain only one person each; both plan on a steady increase in their work load over the next eight months such that by the end of that period they expect to produce eight times the work output they do now. Both plan to add a person per month to keep up.
How successful with they be? The next two charts show all. They will both be dismal failures.
[[[ A -- Half hour meetings ]]]
[[[ B -- Hour Meetings ]]]
=================================================================
Effects on Meeting Time of Adding Members -- Group A
Computed Time used up in one-on-one meetings with average length of .50 hours.
People Total Group Meeting Time added Total Group Added Meeting Work Time Total Group
in group Man-Hours Time Coef by new person Meeting Time Time added Work Time
1 8 0 .00 8.00
2 16 1 8 1.00 1.00 7.00 15.00
3 24 3 8 3.00 2.00 6.00 21.00
4 32 6 8 6.00 3.00 5.00 26.00
5 40 10 8 10.00 4.00 4.00 30.00
6 48 15 8 15.00 5.00 3.00 33.00
7 56 21 8 21.00 6.00 2.00 35.00
8 64 28 8 28.00 7.00 1.00 36.00
9 72 36 8 36.00 8.00 .00 36.00
Effects on Meeting Time of Adding Members -- Group B
Computed Time used up in one-on-one meetings with average length of 1.0 hours.
People Total Group Meeting Time added Total Group Added Meeting Work Time Total Group
in group Man-Hours Time Coef by new person Meeting Time Time added Work Time
1 8 0 .00 8.00
2 16 1 8 2.00 2.00 6.00 14.00
3 24 3 8 6.00 4.00 4.00 18.00
4 32 6 8 12.00 6.00 2.00 20.00
5 40 10 8 20.00 8.00 .00 20.00
6 48 15 8 30.00 10.00 -2.00 18.00
7 56 21 8 42.00 12.00 -4.00 14.00
8 64 28 8 56.00 14.00 -6.00 8.00
9 72 36 8 72.00 16.00 -8.00 .00
Formulas used for the Calculations
| A || B || C || D || E || F || G || H |
1 Effects on Meeting Time of Adding Members
2 Computed Time used up in one-on-one meetings .5
3 People Total Group Meeting Time added Total Group Added Meeting Work Time Total Group
4 in group Man-Hours Time Coef by new person Meeting Time Time added Work Time
5 1 A5*8 0 C5*F2*2 8
6 A5+1 A6*8 1 8 C6*F2*2 E6-E5 D6-F6 H5+G6
7 A6+1 A7*8 3 8 C7*F2*2 E7-E6 D7-F7 H6+G7
8 A7+1 A8*8 6 8 C8*F2*2 E8-E7 D8-F8 H7+G8
9 A8+1 A9*8 10 8 C9*F2*2 E9-E8 D9-F9 H8+G9
10 A9+1 A10*8 15 8 C10*F2*2 E10-E9 D10-F10 H9+G10
11 A10+1 A11*8 21 8 C11*F2*2 E11-E10 D11-F11 H10+G11
12 A11+1 A12*8 28 8 C12*F2*2 E12-E11 D12-F12 H11+G12
13 A12+1 A13*8 36 8 C13*F2*2 E13-E12 D13-F13 H12+G13
14
=================================================================
Both groups have added 64 man hours to the payroll. But Group A has added only 36 hours of work time to the group--the rest of the hours are tied up in added meetings to keep up coordination between all the new people.
Group B, which communicates even more than A, suffers dismally under the new load of people. In fact, Group B has actually added so many people that work time is being reduced with each new member. Work time peaked with the fourth man added. In fact, by the time the eighth man is added there is barely enough time in the day to hold all the meetings!
In both sorting and people interactions, we have developed ways to reduce, but not eliminate, this problem. Developing a new sorting algorithm can reduce a sort time. Developing a new social structure, such as the hierarchical organization, can reduce interaction time.
The following table shows the reduction in interactions resulting from establishing a hierarchical organization where each member has no more than two subordinates, and each subordinate talks only to his or her superior. This structure is not often used for people (they like to have more subordinates per leader), but it is often used in computer processes because it represents a structure that is very efficent when one wants to sort through large quantities of information. I use it here because it lends itself to easy analysis.
=================================================================
Number of Contacts between people in a Hierarchical group
People in group: 1 2 3 4 5 6 7 8 9
Contacts / Time: 0 1 2 3 4 5 6 7 8
=================================================================
[[[ 6 - Communications links in a hierarchy ]]]
Now remember that in this organizational structure, people only talk to their direct superiors or subordinates. In this structure, the number of inter-person contacts is reduced dramatically, so the time spent talking with people is reduced. But the time it takes to pass information from any one member to another organization member is longer, and the message is subject to delays, errors, and distortion. As a result, information is slow-moving and reduced drastically in amount. This kind of structure is useful when decisions need to be made quickly and with limited information. Fortunately for us, lots and lots of decisions can be made this way because hierarchical structures are widely used.
There are other ways to reduce contact time. Having a meeting where lots of people listen to one is an efficiency gainer over having lots and lots of one-on-one meetings with that one person. Having an agenda for the meeting, telephones, secretaries, appointment calanders, etc., are all further improvments.
The challange one faces with any of these reduction of time-spent-on-talking strategies is getting the right information transmitted to the right people quickly, accurately, and in context, so there can be a proper response to it. A classic example of this problem shows up all the time in the Armed Forces. Here is one variant on the problem.
A general tells his colonel he wants a helicopter ready for an important surprise inspection trip at 10:00 tomorrow. The colonel tells the major he wants the chopper ready at 09:30 so he can have chance to make sure it is ready for the general. The major tells the captain to have it ready at 09:00 to make sure it's there in time for the colonel to inspect, and on down the line. So, the next day the helicopter shows up at 07:30, and then burns up fuel for two and a half hours while the various inspectors inspect and make sure that the chopper crew are well prepared. Finally the general shows up on schedule at 10:00, but he can't leave because the chopper is now out of fuel.
In the process of communicating information down the chain of command, important information had been lost. The trip was supposed to be a surprise; and because of that none of the people in the chain had communicated fully with the people below them. Context was lost, and with it vital coordination.
Reducing interactions reduces information flow. When information flow is reduced, response time to new information is increased and the probability of error increases. This reduced response time shows up as operational rigidity. This is why bigger organizations seem to move slower than smaller ones.
How much time do you spend in communicating with each person in your group?
How much more time will be devoted to communications when the next person is added to your group? Conversely, how much work time will really be lost by losing a person? Would you be better off adding a person or splitting a group to gain productivity?
Here is one way to help find out.
We are going to analyze just how much of a group's time goes to talking instead of working. Then, presuming the next addition to the group needs to talk as much as existing members, and presuming the group is fully networked (everyone talks to everyone else), how much of the new bigger group's time will be spent in meetings and how much will be available for work? Here are the steps.
First, define your group. A group is people you talk with on a regular and frequent basis. It may be your department or section or whatever. It is probably less than a dozen people.
Second, keep a log for one week of the total time you spend meeting with others in the group (just with others in the group. Meetings with outsiders don't count. However, if you have a meeting with some insiders and someoutsiders, count it.) Have the others in your group keep similar logs too.
Add up the total time spent meeting all members of the group for the week. Divide it by the number of people in the group, and the number of days in the week. This will give you the average meeting time per person in your group.
Total_Time_Group_Members_Spend_meeting_with_Group_Members Average Time
Number of People in Group x Number of working days = ays = " #>Meeting_____
Person
Person" #>
For purposes of this discussion, any time you don't spend meeting with members of your group will be called work time. Work time is defined as non-meeting time--regardless of what you do with it.
Now, consult where you are on the following chart based on the number of people in your group (column 1).
===========================================================================
Effects on Meeting Time of Adding Members
Zero Gain Meeting Time and Meeting Time Coefficients
---------------------------------------------------------------------------
Formulas:
Meeting Time = Meeting Time Coef * Average Meeting Time
Added Work Time = Added Man-hours - Increase in Group Meeting Time
Definitions:
MEETING TIME COEFFICIENT -- This is the multiplying number used to determine how much meeting time is affected by group size.
ZERO WORK GAIN -- This is the maximum average time a person in a group of the specified size can meet with all the others in the group before adding another person will cost the group work time, not add it.
---------------------------------------------------------------------------
Zero Work Gain
People Meeting Meeting Time
in group Time Coef (minutes/day)
1 0
2 1 240
3 3 80
4 6 40
5 10 24
6 15 16
7 21 11
8 28 9
9 36 7
10 45 5
11 55 4
12 66 4
13 78 3
14 91 3
15 105 2
16 120 2
17 136 2
18 153 2
19 171 1
20 190 1
===========================================================================
===========_Effects on Meeting Time of Adding Members Zero Gain Meeting Time and Meeting Time Coefficients---------------------------------------------------------------------------Formulas: Meeting Time = Meeting Time Coef * Average Meeting Time Added Work Time = Added Man-hours - Increase in Group Meeting Time Definitions: MEETING TIME COEFFICIENT -- This is the multiplying number used to determine how much meeting time is affected by group size. ZERO WORK GAIN -- This is the maximum average time a person in a group of the specified size can meet with all the others in the group before adding another person will cost the group work time, not add it.--------------------------------------------------------------------------- Zero Work Gain People Meeting Meeting Time in group Time Coef (minutes/day) 1 0 2 1 240 3 3 80 4 6 40 5 10 24 6 15 16 7 21 11 8 28 9 9 36 7 10 45 5 11 55 4 12 66 4 13 78 3 14 91 3 15 105 2 16 120 2 17 136 2 18 153 2 19 171 1 20 190 1 ===========================================================================" #>
If you are adding a person to your group, multiply the selected coefficent (column 2) by the average meeting time per person you have already determined from the logs. This will give you the total meeting time the group engages in.
=================================================================
Example:
You have done your logs and determined that the average meeting time in your group is 30 minutes (.5 hours). You have five people in your group. If you add a person, the "working hours" (non-meeting hours) you can expect to gain in the group from adding him or her is figured as follows:
.5 average meeting hours x 15 =
7.5 meeting hours in the new six person group
.5 average meeting hours x 10 =
5.0 meeting hours in the old five person group
8 hours added by new person - 2.5 added meeting hours =
5.5 hours added to group's working time
The moral:
If you bring on a person into this situation expecting them to contribute eight additional hours of "work" to your group, you will be disappointed with the results.
=================================================================
It is this grim reality of fully networked communication that keeps groups small and forces hierarchy upon us. The last column, the Zero Gain column (column 3), shows the maximum average meeting time per person you can have in a group of a particular size without actually loosing "work" time if you add a person to the group!
The Meeting Time Coef column (column 2) is useful if you are considering splitting a group. Take your average meeting time, and multiply it by the number of people in each of the new groups. You will have how much time you can expect them to meet.
=================================================================
Example:
Instead of just adding your next person to the group, you decide the group can be split into two groups at the same time. You now have two three person groups. How much "working" time will each group have versus what they would have had as one six person group?
Here is how it is figured:
Six person group
.5 average meeting hours x 15 x 1 group= 7.5 meeting hours
Two three person groups
.5 average meeting hours x 3 x 2 groups= 3 meeting hours
The moral:
When smaller groups can address a task, less time will be taken up with internal meetings. If a job can be partitioned into smaller tasks, then smaller groups will be able to spend less time talking about the problem and more time attacking it.
=================================================================
We all live in a day with a fixed length, so we must be choosy about how we allocate it. One of the ways we allocate our time is by using it to talk to other people.
We also like the benefits of having large organizations. It allows us to concentrate resources on worthwhile goals. However, when we work as part of a large organization, we need to introduce communications limiting strategies or we spend all our time talking and get nothing done.
The amount of time spent communicating in a group will go up in proportion to the combinations of the people involved--which is roughly the square of the people involved--unless communications limiting strategies are employed.
One of these communications limiting strategies is organizing a heirarchical social structure where a person talks only to superiors and subordinates. The benefits of this are a dramatic reduction in the amount of time spend talking with other people--time goes up in direct proportion to the people involved instead of the square. The disadvantage is that as a result of limited talking, information flows slowly, and often inaccurately.
These are the theoretical problems you and I face when trying to organize people. Next I will discuss how the CIA gets involved in all of this, and what we can do about it.
Now, how does the CIA fit into all of this?
The CIA is one of the most striking examples of this paradox of size slowing information flow. Here is a large, hierarchical, bureaucratic organization in the business of quickly collecting, considering, and disseminating information--secret information. They have amassed lots of resource behind their mission: which is to collect secret intelligence information and to process it quickly into a form that is readily available and easily understood by our heads of state. But, because it is big, and hierarchical, and dealing with "secret" information that can be released only on a "need-to-know" basis, the rate of information flow in the organization is probably quite glacial. In other words, the kind of job it is trying to do is not suited to the kind of organization it is.
This is why we see the paradox of an information gathering organization that can't seem to provide information except with controlled leaks and scheduled press releases.
For example, how much did the CIA tell us about Korean Air Flight 007 when it was happening? How much did they tell us about Cherbonyl during the first week? Both of these events have in common that they occurred suddenly and were not planned. In trying to process information about them, the CIA was fighting its own organizational structure in trying to process information just as much as they were fighting any enemy's attempt to conceal it from them.
Exercising control over information flow based on the need-to-know philosophy might best be described as "supra-intelligent information flow": information flows only when both the knower and the knowee agree that the knower needs to know. This is in contrast to "reflexive information flow"--the kind most of us are used to--where information is freely acquired merely by asking for it.
Now, supra-intelligent information flow may sound really impressive (and, By Golly, if the CIA is using it, it should be!), but it just means that a decision must be made (and time taken to make the decision) before information can flow.
Given the inherent organizational problems the CIA faces because it uses hierarchy and supra-intelligent information flow, it's not surprising the number of intelligence gaffes we see. It's more amazing that the organization can function at all! It does so only on the strength of it's clear slow-changing mission goals, excellent management, and well-organized procedures.
Let's look at a simple example of what these clear mission goals, excellent management, and well-organized procedures must overcome in the way of inherent problems when one is trying to deal with information in a heirarchical, need-to-know way.
Suppose three programmers -- Tom, Dick and Harry -- are attending the NCC (National Computer Conference) together and they decide to go to lunch. Now these three programmers use standard reflexive information flow--the kind we are all used to. Their conversation may go something like this.
Tom: Lunchtime, guys; where do you want to go?
Dick: MacDonald's,
Harry: Wendy's.
Tom: Hey, we're in Las Vegas, let's go somewhere different! How about the buffet at the Hilton?
Dick: Okay.
Harry: Sounds good, let's go.
and off they go.
This was an example of fully-networked, fully-reflexive information flow. Each person could talk to any of the others in the group at any time, and each received information from the others mearly by requesting it. It only took six information exchanges to decide where to go. Note that in those exchanges alternatives were suggested, then full consensus was reached, and the decision made.
Meanwhile, on the very same convention floor, there are three CIA agents -- Larry, Curly and their supervisor Moe. They are looking for foreign spies stealing high-technology, but, it being lunch time, they know it is time to find spies taking their noonday repast. Being very dedicated CIA agents, they are "gung ho" on hierarchy and supra-intelligent information flow. Lets listen to how they decide where to go for lunch....
Special Agent Larry to Supervisor Moe: Special Agent Supervisor, are you hungry?
Supervisor Moe to Special Agent Larry: Why do you want to know, Special Agent?
Special Agent Larry to Supervisor Moe: It's lunch time and I wanted to know if you wanted to go to lunch.
Supervisor Moe to Special Agent Larry: That's a valid reason, so I'll tell you. Yes, I am, and I would like to go to lunch. Let me check with Special Agent Curly.
Supervisor Moe to Special Agent Curly: Special Agent Curly, are you hungry?
Special Agent Curly to Supervisor Moe: Why do you want to know, Special Agent Supervisor?
Supervisor Moe to Special Agent Curly: It's lunch time and I wanted to know if you wanted to go to lunch.
Special Agent Curly to Supervisor Moe: That's a valid reason, so I'll tell you. Yes, I am, and I would like to go to lunch.
Supervisor Moe to Special Agent Curly: Good, where would you like to go.
Special Agent Curly to Supervisor Moe: Why do you want to know, Special Agent Supervisor?
This CIA communication procedure carried to its conclusion will require 18 information exchanges and one arbitrary administrative decision--Moe will pull rank and decide where everyone will eat. In so doing he dispenses with trying to reach a consensus and saves another 24 or so information exchanges.
See how this compares to the 6 exchanges and full consensus achieved by the programmers? Now you know where the CIA's reputation for running lean and mean comes from -- they're always late for lunch, and they rarely eat where they want to!
=================================================================
The Full Exchanges
Special Agent Larry to Supervisor Moe: Tom: Lunchtime, guys; where do you want to go?
Special Agent Supervisor, are you hungry? Dick: MacDonald's,
Supervisor Moe to Special Agent Larry: Harry: Wendy's.
Why do you want to know, Special Agent? Tom: Hey, we're in Las Vegas, let's go somewhere
different. How about the buffet at the Hilton?
Special Agent Larry to Supervisor Moe:
Dick: Okay.
It's lunch time and I wanted to know if you wanted to go to lunch. Harry: Sounds good, let's go.
Supervisor Moe to Special Agent Larry:
That's a valid reason, so I'll tell you. Yes, I am, and I would like to go to lunch. Let me check with Special Agent Curly.
Supervisor Moe to Special Agent Curly:
Special Agent Curly, are you hungry?
Special Agent Curly to Supervisor Moe:
Why do you want to know, Special Agent Supervisor?
Supervisor Moe to Special Agent Curly:
It's lunch time and I wanted to know if you wanted to go to lunch.
Special Agent Curly to Supervisor Moe:
That's a valid reason, so I'll tell you. Yes, I am, and I would like to go to lunch.
Supervisor Moe to Special Agent Curly:
Good, where would you like to go.
Special Agent Curly to Supervisor Moe:
Why do you want to know, Special Agent Supervisor?
Moe to Curly
So we can decide where to go for lunch.
Curly to Moe
That's a valid reason, so I'll tell you. I'd like to go to Mac Donalds.
Moe to Larry
Where would you like to eat, Larry?
Larry to Moe
Why do you want to know?
Moe to Larry
So we can decide where to go eat.
Larry to Moe
That's a valid reason, so I'll tell you. I'd like to go to Wendy's.
Moe to Larry
Thank you for your input, Larry. We will be going to the buffet at the Hilton.
Moe to Curly
Thank you for your advice, Curly. We will be going to the buffet at the Hilton.
=================================================================
Keep in mind that as both groups get bigger, the communication process grows more and more unwieldy, but it gets unwieldy faster for the fully networked programmers. If nine programmers are trying to decide where to eat then 72 information exchanges are required -- provided a debate doesn't start about where to go. If nine CIA agents try to decide, they'll also require 72 exchanges. And, since they're still lean and mean, there'll be no debate. (The increase in exchanges caused by supra-intelligence requirements is offset by the reduction in exchanges due to hierarchical organization -- where the members only talk to superiors and subordinates.)
So, what have we learned about the CIA without using a single spy? We have learned that because the CIA is a big organization, hierarchically organized, and operating with supra-intelligent information flow, it will move information very slowly. It operates efficiently only out of the habits inculcated by good management, clear goals, and well-organized procedures. We can expect it to respond very quickly and efficiently to known threats because it is large and has access to large resources to meet the challange. We cannot, however, expect it to respond quickly to the unexpected. And since it cannot respond well to the unexpected, we can expect there will always be intelligence gaffes, and that the myth of the CIA always being in the right place at the right time with the right people will, also, always be a myth.
What does this mean to us as people in organizations? It means that if we are dealing with a large organization, we can expect it to respond very slowly to new information or situations. It may respond very well to situations that have been brewing for a long time, and that may look like a quick response--such as our attack on Libya. But in reality we can expect few surprises coming from within large organizations. Not none, but few. And that if events around us move quickly and in unexpected ways, large organizations will not be able to respond appropriately or rapidly. Cherbonyl is a good example of that--no one was well-prepared for that, and it took days for even the Russians to find out what had happened.
Because of this, we can expect small organizations to remain viable. And as our society goes through faster and faster changes, we can expect small organizations to become more and more viable, not less and less.
Next I will discuss how the size an organization will affect the kinds of goals it can expect to attain, and what alternatives we have to hierarchical organization.
Any hierarchy, by its very nature, restricts information flow, even without a need-to-know restriction. Subordinates decide what is important for a supervisor to know. Supervisors decide what is important for the subordinates to know. And both remove information from its context as they pass it on, which makes it prone to misinterpretation. This problem is further compounded by any overt errors that occur in the transmission.
Given all these problems, why deal with hierarchy at all? The great advantage of hierarchy is its ability to marshall big resources behind a goal. This is something small groups can't do. Its great weakness is trying to keep the goal in sight. If the goal doesn't move, like building a pyramid on the desert, then hierarchy works famously. But what if the goal changes?
If organizational goals do change, then smaller structures with more information cross-ties work better. With a smaller group and more information cross-ties, information flows faster, so the group will be more responsive to changes. This is why armies have squads, regiments, divisions, brigades, etc. This is one of the reasons why departments, decentralization, and profit center organizational structures work well in big companies.
As we have disucssed previously, if a big group attempts full networking all the individuals' time will be spent talking and none doing. Now, can you think of a good example of a place where people are working when they spend all their time talking and get nothing done?
Full networking is a good model of what happens at a conference or convention. A big part, if not a majority, of the time spent at a conference or convention is spent finding and meeting with small groups of people, or listening to one person say something to a large group. Which leads us to the next way of dealing with large groups of people.
=================================================================
What is Robert's Rules of Order?
When large groups of unorganized people must be delt with simultaneously, we call upon Robert's Rules of Order. Which, if you look at it from a data processing point of view is nothing more than an Input/Output processing protocol. It divides the inputs into various catagories (motions, amendments, calls, points of order, etc.), details their priorities (a motion to ajourn, for instance, is top priority--which is interesting since this is the equivalent of a Reset command in a computer I/O processor. And the Reset command is also a top priority command.), then explains how to process them (vote on them, put them to commitee, table them, etc.), and what to do with the outputs (make a report, note in the minutes, etc.). In fact, I'm strongly tempted to do a formal comparison of these two I/O processors one of these days.
=================================================================
Organization size should be appropriate to the goal being sought and the resources needed to accomplish the goal. If the goal is fixed and easily defined, and the resources required are substantial, then the size of the organization can be large. If the goal is a moving target, or a bit fuzzy, then the organization should be smaller because it will be more flexible and communications about "what the goal is today" will be easier to disseminate. If the goal is moving, but requires large resources to accomplish, then you have a true organizational challange. The most efficient way to solve the problem may be to go back and reexamine the goals to see if they can be changed. If not, then expect that it will require exceptional people for the organization to meet its goals.
There are other ways of organizing besides hierarchy, but they tend to serve special, non day-to-day purposes. Conventions are an organization designed to let lots of people meet and talk in small groups. Meetings and congresses are organizations designed for lots of people to process ideas through a central point. When Robert's Rules of Order are invoked for a meeting then the whole meeting operates very much like an I/O processor in a computer system.
Even with excellent management, one of the biggest problems a large hierarchical organization of the CIA's sort faces is dealing with intuition. When dealing with intuitive processes, who knows all he or she has a need-to-know? Carried to its extreme, hierarchy and supra-intelligent communications (besides taking a long time) make intuitive connections all but impossible to uncover--there is no need-to-know until the need-to-know has been proven--so intuitive information won't flow. This flow restriction is in addition to other limiting effects of hierarchy information flow--such as it may not be clear who in the organization would know in the first place.
Even when the need-to-know philosophy is not carried to its extreme; even when it is just jawboned; as the Department of Defense is currently doing with it's flap over American universities leaking vital secrets to the Soviet Union by giving open conferences on science and high technology, and the Pentagon is doing when it complains about businesses selling high-technology to our enemies--it is very hard on intuition. Intuition is a right brain function, and the right-brain functions are both very inarticulate and very impatient.
=================================================================
FYI
To read an example of a classic confrontation between need-to-know and reflexive-holistic attitudes towards high-tech sales to foreign countries, try the case study in Rock and Simmonds's International Business and Multinational Enterprises 3rd Ed. Technology Transfer: The Cyber 76chnology Transfer: The Cyber 76" pages 632-636.
=================================================================
The more we are restricted to need-to-know for our own information needs, the less advantage we have over other societies and organizations where need-to-know is the SOP (Standard Operating Procedure). As we enter the information age, the right-brain functions of intuition and holistic thinking are becoming more and more important. These are the areas where we in our advanced society have the resources to excel: intuition is useless to the person who can't understand a problem, is asked only to dig a ditch, or is watching his family die of malaria. But to the person who is well-educated, can direct thousands of horsepowers, dozens of robots, and millions of dollars, these functions are the competitive edge--the four-wheel drive that keeps an organization from getting stuck in the mire of mediocrity. These are the areas where all the crazy thinking of the 60's "me" generation can finally bear useful fruit.
Just as important, need-to-know attitudes are self-defeating. The more we try to restrict information flow to our enemies, the more we succeed in restricting it to ourselves--we cut off our nose to spite our face. Our society's needs for information are a million times larger than any spies' needs, and the benefits we can make of that information are a billion times greater than the damage any spy can do irregardless of whether the spy comes from a competing company or a competing nation.
Our society has spent the last two centuries structuring itself to promote and take advantage of free information flow. A society not trained in free information flow cannot, I repeat, CANNOT, make use of information as quickly or efficiently as we can. Look how long it took the Russians to report on the Cherbonyl Disaster. It wasn't so much because they were intentionally dragging their feet as it was that their own organizations were and are not equipped to handle the exceptional information flow that a major disaster like this represents. And the Russians are an advanced country populated with intelligent people. What threat are Third World countries like India or China or Ghana where not only is there no heritage of free information flow, but no big education resource to fall back on either? Why lower ourselves to that level of existence?
There will be spectacular exceptions, of course, to this basic rule that we can do more with our own information than any one else can. Execptions such as Third World countries making atomic bombs or the Russians getting into space first. But the exceptions will be in areas where the non-free society has focused its goals so hierarchy can work effectively--the exceptions will remain few and far between. These are examples of what a writer in Technology Review called Technological Potemkin Villages.
Potemkin's villages were shams. They were put up by Potemkin, a minister in Czarist Russia, to fool his queen, Cathrine the Great, on a tour of land Potemkin was supposed to have colonized. The queen bought the ruse; Potemkin became a legend; and the Russians learned a new way to overcome problems with their hierarchy--they published reports that the problems didn't exist.
Technological Potemkin Villages are highly publisized triumphs of Second or Third World technology that are quite isolated in nature and barren of further implications. Such things as building jet fighters that are designed to fly as fast as Mach 3 -- but if they do so for more than a few seconds, the plexiglass windows of the cockpit melt.
If we remain open, we can have our cake and eat it too, and have lots more besides. Our technological advances feed upon themselves -- if we let them. We just need to keep in mind that one of our essential long term social goals has to be free information flow.
So, is supra-intelligent information flow a part of your organization? Do you use it? Does your organization abuse it? Here is one quick check, answer this:
Do you know what your organization's mission statement is? How did you find out?
Now the mission statement is a piece of information that in theory everyone in the organization should know if they are going to help the organization towards its goal. But, based on the efforts of many of my compatriots at MBA school, many organizations consider this classified information. In one case, it was only the secretary that had the information filed that considered it classified, but her VP didn't know that. Does that constitute supra-information flow? It sure does! It may be involuntary on the part of the organization, but the results are the same. Suppose you had found out about the mission statement "through the grapevine". Could you be sure you got it right? The point of this is that supra-intelligent, need-to-know type information restrictions can creep into an organization without any direct action by managment to bring them on. They can spring up spontaneously out of ignorence unless there are clear procedures to the contrary. They are something you must always be vigilant for in your organization.
So, there you have it.
"The cures to all what ails your organizations, my friends. All here in this little bottle and yours for only a dollar."
Well, not realy, but at least some insights into why organizations are, and whether they're as close to right as they should be.
Let's review:
First, look at an organization's size, then at its goals. Then ask: Can it be made smaller? Can it be divided logically into smaller units that will maintain better information flow within their areas of responsibility? If either of these can be done, then the strangling effects of limited information flow that gets worse with the size of the organization will be relieved.
Second, do you deal in supra-intelligent information flow? Do you really mean to? Can you do something about it? Can you make and enforce a policy that anyone gets any information they want, unless specifically excluded from having it?
An organization must take an active effort to make sure information gets to the right places if it wishes to keep the information restricting effects of big organizations from gaining the upper hand. Good computer systems help, good clear policy helps even more. This is especially true in fast moving organizations that have to meet the demands of fast changing markets or fast changing technologies.
Remember that the first thing to suffer from restricted information flow is intuition. If you are looking for innovations and breakthroughs, you need to have (or look at) small organizations with no information restrictions.
These are the challanges I put to you:
And don't hog communications lines you don't need. If two people have a problem to solve, let them solve it themselves. But if you're their supervisor, make sure they've solved it.
If you do these things, then you'll be making your own life easier and life more aggreeable to all around you. If things aren't right, you'll know where to look to make a change.