The garbage can model

In the classical managerial imagination, decision-making is seen as a rational process. A manager identifies a problem, gathers information, compares options and then chooses the most appropriate solution. This representation continues to permeate many discourses on leadership, strategy and governance. However, concrete observation of organizations often reveals a much more uncertain reality.

Decisions sometimes emerge without any real prior analysis, certain solutions precede the problems they are supposed to solve, projects emerge because funding already exists, and meetings produce more displacement of problems than actual resolutions.

It is precisely this reality that the Garbage Can Model attempts to describe. Developed by Michael D. Cohen, James G. March and Johan P. Olsen in a classic article published in 1972, this model represents a major breakthrough in the history of organization theory. Its aim is not to show how organizations should decide, but how they actually do, in contexts marked by ambiguity, complexity and fluidity of commitment.

From American universities to "organized anarchies"

The model emerged in a particular intellectual context. At the end of the 1960s, management science and organization theory were still strongly dominated by the idea of rationality. Although Herbert Simon had already introduced the notion of bounded rationality, the organization was still largely thought of as a coherent system oriented towards identifiable objectives.

At the time, Cohen, March and Olsen were working on American universities. But they resisted conventional models. Objectives are multiple and sometimes contradictory; technologies of action are poorly understood; participants are constantly moving in and out of decision-making processes. The authors describe these systems as "organized anarchies" .

Three characteristics define these organizations:

• problematic or ambiguous preferences ;
• fuzzy technologies;
• fluid stakeholder participation.

In other words, organizations don't always know exactly what they want, don't fully understand how they produce their results, and see their participants become involved intermittently.

In this context, decision-making no longer appears to be a linear process. The authors propose a provocative metaphor: decision-making resembles a dustbin in which problems, solutions, participants and opportunities for choice circulate simultaneously. These different flows come together more or less randomly. A decision is reached when they temporarily coincide.

One of the model's strongest intuitions is that solutions often exist before problems. Actors already have ideas, tools or projects that they then seek to link to a decision-making opportunity. The decision then becomes less a rational resolution than a contingent coupling between available elements.

This approach represents a profound critique of the heroic visions of the strategic leader. Organizations appear more like systems traversed by multiple temporalities, fragmented interests and competing logics.

A fruitful but controversial model

The model's success was considerable. It quickly spread beyond the academic field, to be used in the analysis of public policies, administrations, large companies and territorial organizations. Its influence is particularly strong in work on public action, notably by John Kingdon, who uses it to analyze political agenda-setting.

The garbage model has several strengths.

• Firstly, it provides a realistic description of organizational environments marked by uncertainty. It explains why some important decisions go unheeded, while secondary problems mobilize considerable energy.
  
  
• It also sheds light on the phenomena of managerial fads: certain solutions circulate independently of the real needs of organizations, and then look for problems to which they can be applied.
  
  
• The model also shows that decisions sometimes serve more to maintain organizational activity than to solve problems. Meetings are organized to produce symbolic coordination, decisions are taken to give a sense of action, and mechanisms are created to absorb tensions without necessarily resolving them.

However, several researchers have pointed out the limits of this model.

• Some criticize it for being excessively chaotic. Gary Mucciaroni, for example, believes that the model tends to underestimate the power relationships, political strategies and forms of real rationality that nevertheless structure organizations. Decisions are not always as random as he suggests.
  
  
• Other criticisms concern its descriptive character. The model explains certain ambiguous situations relatively well, but offers few levers for action to improve decision-making processes. It describes disorder without always indicating how to regulate it.

Erhard Friedberg, however, qualifies this criticism. For him, the main merit of the model is that it reminds us that organizations never operate according to perfect rationality. It forces us to recognize the real depth of human interactions, conflicts of interpretation and constant adjustments.

Over time, the model has also been enriched by computer simulations and computational approaches. Fioretti and Lomi show, for example, that apparently inefficient behaviors such as postponing decisions or transferring problems to other players can paradoxically contribute to the overall stability of certain complex organizations.

The garbage model thus ceases to be merely an ironic critique of bureaucracies. It becomes a theory of complex systems confronted with multiple flows of information, actors and temporalities.

Astonishingly relevant in the age of AI and fluid organizations

More than fifty years after it was first formulated, the garbage can model has taken on new relevance. Contemporary organizations often resemble more the "organized anarchies " described by March and his colleagues than the stable hierarchical structures imagined by classical management.

The proliferation of cross-functional projects, digital platforms, hybrid groups and information-saturated environments is accentuating the fragmentation of decision-making. Managers are confronted with an abundance of data, accelerated temporalities and mobile or intermittent participants. Decisions frequently emerge from opportunistic crossroads between available technologies, media urgency, budget constraints and network effects.

Digital transformations also reinforce the circulation of "solutions seeking problems". Artificial intelligence is a striking example. Many organizations are now deploying AI tools before they have even clarified the precise needs they are designed to meet.

Indeed, researchers are beginning to explicitly revisit the garbage model in the age of AI. Joshi and his colleagues show that data scientists, algorithmic tools and massive data flows are helping to make contemporary decision-making dynamics even more complex.

The model is nevertheless confronted by other recent approaches. Theories of distributed decision-making, collective intelligence and complex adaptive systems place greater emphasis on the emergent self-organizing capacities of collectives. Where the garbage model sees mainly disorder, these approaches sometimes observe forms of implicit coordination.

• Models inspired by the sciences of complexity, notably those of Karl Weick or Ralph Stacey, consider that organizations continually produce meaning through their interactions. Decisions are not simply random encounters between independent flows, but emergent collective constructs.
  
  
• Similarly, contemporary approaches to the learning organization, inspired in particular by Peter Senge, assume that a collective can develop reflexive capacities enabling it to reduce certain forms of decision-making chaos. Where the garbage model insists on structural ambiguity, learning organizations seek to develop spaces for dialogue, clarification and coordination of meaning.

For all that, the model retains considerable critical force. It reminds us that every organization is traversed by heterogeneous logics, disjointed temporalities and partial rationalities. It warns against illusions of total control, particularly strong in contemporary discourse on data, algorithmization and indicator-based management.

The garbage can model remains valuable, not because it provides a decision-making method, but because it introduces a form of organizational humility. It reminds us that deciding is not simply a matter of applying abstract rationality. Deciding also means coping with disorder, coincidences, conflicts of attention and the unpredictable dynamics inherent in human collectives.

References

Cohen, M. D., March, J. G., & Olsen, J. P. (1972). A garbage can model of organizational choice. Administrative Science Quarterly, 17(1), 1-25.

Fioretti, G., & Lomi, A. (2010). Passing the buck in the garbage can model of organizational choice. Computational and Mathematical Organization Theory, 16(2), 113-143.

Friedberg, E. (1997). Organization theory and the question of organized anarchy. Paris : PUF.

Kingdon, J. W. (1984). Agendas, alternatives, and public policies. Boston: Little, Brown.

March, J. G. (1988). Décisions et organisations. Paris: Éditions d'Organisation.

Simon, H. A. (1991). Bounded rationality and organizational learning. Organization Science, 2(1), 125-134.

Weick, K. E. (1995). Sensemaking in organizations. Thousand Oaks : Sage.