A Garbage Can Model of Organizational Choice | M. Cohen, J. March
This work summarizes one of the most important
papers written by Cohen and March. It’s made by a
student for students, therefore probably it contains
both typos, both conceptual errors.
It does not substitute the reading and understanding
of the original paper.
GIULIA CORSI
1
A Garbage Can Model of Organizational Choice | M. Cohen, J. March
A GARBAGE CAN MODEL OF ORGANIZATIONAL CHOICE
Michael Cohen, James March
ABSTRACT
Organized anarchies: (ex: University)
Organizations characterized by three general properties:
1. Problematic preferences = It is difficult to impute a set of preferences to a decision situation that satisfy consistency requirements for
a theory of choice. It can be better described as a collection of ideas than as a coherent structure: it discovers preferences by action
then basing on the basis of preferences.
2. Unclear technology = Although the organization manages to produce, its own processes are not understood by its members: it works
on simple trial&error procedures, learning from past experience without a precise structure/method.
3. Fluid participation = Participants vary in the amount of time and effort they devote to different domains; as a result the boundaries of
the organization are uncertain.
In order to build theories of organization that accommodate the organized anarchies model two phenomena must be investigate:
a) Manner in which organizations make choices without shared goals
b) Way members of an organization are activated (how occasional members become active and how attention is directed toward -or
away from- a decision)
Other smaller concepts needed
c) Normative theory of intelligent decision making under ambiguous circumstances (which is the meaning of intelligence that does not
depend on known goals?)
d) Normative theory of attention (who is attending to what?)
e) It is required a revised theory of management, since the contemporary ones use mechanisms to for coordination and control that
assume the existence of well defined goals and technology, without which many axioms collapse.
In this paper a garbage can model for describing decision making within organized anarchies is developed, and the impact of some aspects of
organizational structure on the process of choice within such a model are examined.
THE BASIC IDEAS
Organizations can often be viewed conveniently as vehicles for solving well-defined problems and provide sets of procedures through which
participants arrive at an interpretation of what they are doing and what they have done while in the process of doing it. From this point of view,
an organization is a collection of choices looking for problems, solutions looking for issues, decision makers looking for work. One can view a
choice opportunity as a garbage can into which various kind of problems and solutions are dumped by participants as they are generated. The
mix of garbage in a single can depends on the mix of cans available and on the labels attached to alternative cans. It may be convenient to
imagine that choice opportunities lead first to the generation of decision alternatives, than to an examination of their consequences, then to an
evaluation of those consequences in terms of objectives, and finally to a decision. But this type of model is often a poor description of what
actually happens. In the garbage can model, on the other hand, a decision is an outcome of several relatively independent streams within an
organization.
Attention is limited here to interrelations among four such streams:
a) Problems: They might arise over issues of lifestyle, family, work, careers, jobs, money, ideology, etc…
b) Solutions: A solution is somebody’s product. Despite the dictum that you cannot find the answer until you have formulated the
question well, you often do not know what the question is in organizational problem solving, until you know the answer.
c) Participants: Participants come and go, since every entrance is an exit somewhere else.
d) Choice opportunities: Occasions where an organization is expected to produce behavior that can be called a decision.
Although not independent of each other, each of the streams can be viewed as independent and exogenous to the system. We will concentrated
on examining the consequences of different rates and patterns in flows in each of streams and procedures for relating them.
THE GARBAGE CAN
A simple simulation garbage can model can be specified in terms of this four streams, each considered as a function of time f(t).
a) Stream of choices: We have a number m of choices assumed, each characterized by an entry time (calendar at which the choice is
activated for decision) and a list of participants eligible for that choice.
b) Stream of problems: Number w of problems assumed, each with an entry time (calendar at which the problem becomes visible), an
energy requirement to solve a choice to which the problem is attached and an access structure (list of choices to which the problem
has access).
c) Rate of flow of solutions: Focus is on the rate at which solutions are flowing into the system; since the organization is mutable, we
assume that different energies are required to solve the same problem at different times. We define a solution coefficient ranging
between 0 and 1 (potential energy to determine the output).
d) Stream of energy from participants: There is some number v of participants each characterized by a time series of energy available
for organizational decision making.
Two mapping structures are needed to describe this model:

Decision structure: Mapping of choices onto decision makers described by a matrix D in which dij is 1 if the ith participant is eligible to
participate in the making of the jth choice. Otherwise, dij is 0.
GIULIA CORSI
2
A Garbage Can Model of Organizational Choice | M. Cohen, J. March

Access structure: Mapping of problems onto choices described by a matrix A in which a ij is 1 if the ith choice is accessible to the jth
problem. Otherwise, aij is 0.
In order to connect these variables three assumptions are specified:
1. Additivity of energy requirements assumption: Each choice requires as much energy as the sum of all required energy of the several
attached problems; the energy devoted is the sum of the energies of decision makers attached to that choice in each time period.
2. Energy allocation assumption: The energy of each participant is allocated to no more than one choice per time period, and only
considering choices for which the participant is eligible.
3. Problem allocation assumption: Each problem is allocated to no more than one choice per time period, and only considering choices
accessible for the problem.
These assumption can be modified, but we not pursued the consequences of these modifications here.
ORGANIZATIONAL STRUCTURE
Elements of organizational structure influence outcomes of a garbage can decision process by:

Affecting the time pattern of the arrival of problem choice, solutions or decision makers (changes linkages among the various streams,
affecting eligibility of participant and accessibility of choices for problems in different time periods)

Determining the allocation of energy by potential participants in the decision
The organizational structure changes as a response to actions resulting from a mixture of managerial planning, individual and collective learning,
and imitation.
The paper present some organizational structures such as:

Net Energy Load=Difference between the total energy required to solve all problems and the total effective energy available to the
organization over all time periods. When this is negative, there is enough energy available. We can have different load situations: we
have light load, (ex. -50) moderate load (ex.-25) or heavy load (0 -> energy required = energy available).

Random Entry Times= Reflects the entry times for choices (one choice enter per time period in a certain random order).

Access Structures= The access is represented by the access matrix (A). We can have:
o
Hierarchical access: important problems (rows with low number) have access to many choices and important choices
o

1 1
1 1
1 1
Unsegmented access: any active problem has access to any active choice (matrix full of 1).
(columns with low number) are accessible only to important problems.

1
A1 : 1
1
o
1
1
0
𝐴2 :
0
0
0
1
1
1
0
0
0
1
1
1
1
0
0
1
1
1
1
1
0
1
1
1
1
1
1
1
1
Specialized access: each problem has access to only one choice and choices are accessible to only two problems. 𝐴3: 00
0
0
0
0
1
1
0
0
0
0
0
0
1
1
Decision Structures: same as access structures, we have unsegmented decisions (any decision maker can participate in any |active
choice opportunity), hierarchical decisions (important choices must made by important decision makers), and specialized decision
(each decision maker is associated with a single choice, but each choice has also a single decision maker).
Energy Distribution= We can recognize three types of energy distribution:
o
Important people – less energy: people important in the hierarchical decision structure have less energy (this can have
implications in the motivation within the organization)
o Equal energy: there is no difference among all decision makers respect to energy
o Important people – more energy: Opposite than the first.
SUMMARY STATISTICS
The garbage can model operates under each of the possible organizational structures we have seen to assign problems and decision makers to
choices, to determine the energy required and effective energy applied to choices, to make such choices and resolve problems.
Decision style= decisions are made in three different ways:
1. By resolution: Some choices resolve problems after some period of working on them
2. By oversight (errore): If a choice is activated when problems are attached to other choices and if there is energy available to make
the now choice quickly, it will be made without any attention to existing problems and with minimum time and energy.
3. By flight: In some cases choices are associated with problems (unsuccessfully) for some time until a choice more attractive to the
problems come along. The problems leave the choice, and thus it is possible to make the decision. But the decision resolves no
problems, and they are now attached to a new choice.
Problem activity=Amount of time unresolved problems are actively attached to choice situations
Problem latency=A problem may be active, but not attached to any choice. The problem latency is the amount of time problems spend
activated but not linked to choices.
Decision maker activity=Is the measure of the degree of decision maker activity in the system, which reflects decision maker energy
expenditure, movement and persistence. To measure it, we use as parameters:

The total number of time periods a decision maker is attached to a choice.

The total number of times a decision maker shifts from one choice to another.
GIULIA CORSI
3
A Garbage Can Model of Organizational Choice | M. Cohen, J. March

The amount of energy available and used.

The total energy used on choices in excess of the required to make them at the time they are made.
Decision difficulty= To measure it we use two alternative strategies:

Strategy 1: Consider the total number of choices not made by the end of all the time periods

Strategy 2: Consider the total number of periods that a choice is active, summed over all choices
IMPLICATIONS OF THE MODEL
An analysis of simulations of this model shows eight major properties of the garbage can decision process:
1. Decision making by flight and oversight is a major feature of the process (instead of resolution)
2. The process is sensitive to variations in load: an increase in the net energy load generally increases problem activity, decision maker
activity, decision difficulty, and the use of flight and oversight; choices are likely to take longer and are less likely to resolve problems.
3. There is the tendency of decision makers and problems to track each other through choices.
4. There are important interconnections between problem activity, problem latency and decision time (persistence of choices). It has
been proved that segmentation of access structure tends to reduce the number of unsolved problems active, but increasing the
latency period, whereas segmentation of decision structure decreases problem latency, but increases problem activity and decision
time (each choice is related to a decision maker but it takes more time for him to do it alone).
5. The process is frequently sharply (bruscamente, improvvisamente) interactive.
6. Problems that appear early are more likely to be solved than later ones.
7. Important choices (made by flight and oversight) are likely to resolve problems than unimportant choices (made by resolution).
8. The choice failures that do occur are concentrated among the most important and least important choices; choices of intermediate
importance are virtually always made.
GARBAGE CANS AND UNIVERSITIES
Universities are organizations in which decision situations involve unclear goals, unclear technology, and fluid participants (perfect example of
application for the garbage can model). University decision making frequently does not resolve problems; choices are often made by flight or
oversight. Problems, choices and decision makers arrange and rearrange themselves and in the course of these arrangements the meaning of a
choice can change several times. Problems are often solved, but rarely by the choice to which they were fist attached.
Slack (esubero) = is the difference between the resources of the organization and the combination of demands made on it; it is sensitive to two
major factors: money (and resources in general) and the internal consistency of the demands made on the organization by participants. The
consequences of slack changes in a garbage can decision process can produce changes in the organizational structures adopted in the decision
process.
Adversity=condition in which the slack is reduced.
Considering the resources available (money and human) we can identify four types of colleges and universities, with their relative adopted
structures:
LOAD
ACCESS STRUCTURE
DECISION STRUCTURE
ENERGY DISTRIBUTION
LARGE, RICH
LIGHT
SPECIALIZED
UNSEGMENTED
LESS
LARGE, POOR
MODERATE
HIERARCHICAL
HIERARCHICAL
MORE
SMALL, RICH
LIGHT
UNSEGMENTED
UNSEGMENTED
MORE
SMALL, POOR
MODERATE
SPECIALIZED
SPECIALIZED
EQUAL
Large, rich schools will be characterized by a high degree of problem
latency.
The large, poor schools are in the worst position under adversity: they
have a high level of problem activity, a substantial decision time and a low
level of decision makers mobility.
What happen in adversity conditions? What about exit opportunities? ---->
CONCLUSION
A set of observations made in the study of university organizations as
organized anarchies has been translated into a garbage can model of
decision making, since we have not the conditions to apply classical models
related to the fact that: preferences are problematic, technology is unclear,
participation is fluid. The garbage can process is one in which a mix of
problems, a mix of solutions and a mix of participants move from one choice
in the garbage can to another at any time. Four factors have substantial
effects on the operation of the garbage can process: the organization’s net
energy load, the energy distribution, its decision structure and problem
access structure. It is clear that the garbage can process does not resolve problems well, but it enable choices to be made and problems resolved
even when the organization is plagued with goal ambiguity and conflict, with poorly understood problems.
GIULIA CORSI
4