"Garbage Can" Models: Multiple Stream Theory

Overview

The "garbage can model" emerged as part of a critique of rational and neo-rational models of public administration, such as those of Herbert Simon (Cohen, March, & Olsen, 1972). Rather than portray decision-making in public administration as a matter of rational choice, John Kingdon and other theorists in this school have described it as a process characterized by organizational anarchy. Organizations do not function like computers solving optimization problems. Rather they function like garbage cans into which a mix of problems and possible solutions are poured, with the precise mix determining decision outcomes. The mix reflects how many decision areas are handled by the organization, how people have access to the organization, the decision load of the organization, and the organization's level or resources, time, energy, and attention. This theory has been used particularly by students of public-sector budget decision-making.

Key Concepts and Terms

Multiple streams. Processes in the garbage can (organization) include those associated with problems, politics, and policies (Kingdon, 1984, 1995). The problem stream revolves around agenda-setting processes. The political stream revolves around contention over alternatives and reflects public opinion, interest groups, experts, elections, partisan forces, and legislative, judicial, and executive bodies. The policy stream revolves around defining policy solutions, often describable as much as a process where favorite solutions are looking from appropriate problems as one in which problems lead to solutions (Cohen, March, & Olsen, 1972: 1).
An implication is that administrative decisions cannot be understood in purely rational terms. Rather decisions must be understood in the context of the three process streams which determine the precise mix in the garbage can. This means the analyst must look at how problems coming along in the organization's pipeline percolate to the top of the agenda, and how various players (president, Congress, interest groups, media, policy communities, policy entreprenuers, public opinion, interest factions within the organization) contend over possible solutions in the conflict-ridden whitewater of multiple streams of problems, interests, and options channeling toward policy formation.
- The stream metaphor. Policy streams are organized into channels. Channels are largely separate from one another. When channels merge, whitewater results from the force of unlike channels. Streams are not static, circular, or oscillating, but rather flow toward policy solutions, albeit not in orderly, rational ways. Streams have a visible surface but deep hidden forces and tendencies. Participants are not so much in control of the stream as carried along with it, only partially able to guide their direction.
- The communities metaphor. Multiple streams theorists employ a second metaphor centering on the idea of policy communities, which are specialised but fragmented networks interested in one or more aspects of a policy. Policy entrepreneurs create alliances among policy communities, based on mutual interest and compromise. Policy windows open unpredictably due to shifts in the policy stream, giving special opportunities to policy entrepreneurs to advance their causes based on changed events or changed environment. Policy windows may allow for non-incremental policy changes, even though most of the time incrementalism is the rule.
- Multiple streams in budgeting. Irene Rubin (2006, 2007) has adapted 'garbage can' theory for the setting of federal budgeting, which she interprets as reflecting the convergence of five streams: revenue (taxes, fees, tariffs, etc.), expenditure (what to spend on), process (who influences what to spend on), implementation/execution (how to administer), and balance (how to meet balance requirements). In what she labels "real-time budgeting," budget decisions are made in a non-linear manner, on the fly, in the context of pressures from any of the five streams or some combination converging at the moment. The multiplicity of streams creates a variety of dynamics. For instance, when actors fail to prevail in the expenditure stream they may seek to focus on restructuring the process stream. When decisions in the expenditure stream precede decisions in the revenue stream, they will be less stable. Etc. Similarly, Katherine Willoughby (Melkers and Willoughby, 1998, 2001, 2005; and Thurmaier and Willoughby, 2001) has analyzed budgeting in terms of "multiple rationalities," referencing Rubin's concept of "real-time budgeting" but arguing that performance budgeting would increase the flow of information in agenda-setting streams (she references the agenda-setting model of Kingdon, 1984).
Coupling occurs when multiple streams coincide, as when events force a potential problem solution to the fore, political actors are predisposed by interest toward the solution, and policy communities have defined and achieved relative consensus on the solution. Coupling may be partial or pervasive, and may be brought about by random events or by the concerted action of policy entrepreneurs. Coupling may lead to synergistic creation of new, common policy alternatives. Coupling may greatly increase the likelihood of a given alternative becoming adopted public policy.
- Loose coupling refers to the fact that the intentions of individuals may not be acted upon, and even organizational goals may be only loosly coupled with organizational actions. This means that individual and organizational decision-making may fail to coincide. Various reasons for this exist. Individual and organizational goals may be ill-defined and at variance (problematic preferences). The relation of means to ends may not be well understood (technological indeterminacy). Because the organization is a collection of "garbage cans" which are semi-autonomous, problem-solving may fail to converge (unstable integration).
Blocking is an anti-coupling strategy employed by some agenda generators, especially interest groups, which often are most intensely motivated and activated by agendas they dislike, leading to their undertaking negative blocking efforts. In the stream metaphor, those struggling to guide the interest group raft are more urgently motivated to avoid crashing on mid-stream rocks than they are motivated to plan for advantages which may come from the coupling of streams in the distance, though both occur.

Assumptions

Anarchically individualistic rather than rational decision-making. Problems and solutions are seen as attaching to individuals who enter and exit decision arenas ("garbage cans"). Therefore the mix of individuals who exist at any given time in the garbage can determines what problems rise to the top of the agenda and what solutions are associated with them. Policy outcomes are then the result of anarchic decision-making within the garbage can. Outcomes cannot ordinarily be predicted by rational analysis of what problems are likely to emerge or on the basis of optimal solutions to projected problems. An implication pursued by Padgett (1980) is that organizational decision-making might be simulated using demographic starting points and random, stochastic processes.
- Critique of anarchy. Bendor, Moe, & Shotts (2001: 182) used simulation methods to critique garbage can theory's assumption of anarchy in decision-making: "Decision makers in the verbal [garbage can] theory contront a chaotic world in which they, solutions, and problems dance around one another, meeting by chance in choice arenas. But in the simulation, packs of decision makers - and often problems as well - march in lockstep from arena to arena. And solutions never move at all."
Rationalization. Because of the anarchic and individualistic nature of organizational decision-making, official presentations of orderly policy with respect to a problem generally represent post-hoc rationalizations of what has been decided largely on a non-rational basis.

Illustrative Hypotheses

The more the people involved in the decision process, the larger the number of solutions proposed and the more anarchic the decision process.
The more the turnover of people involved in the decision process, the more anarchic the decision process.
The more problematic the preference structure, the less the convergence of individual and organizational decision preferences.
The more the technological indeterminacy, the less the convergence of individual and organizational decision preferences.
The more the autonomy of formal and informal decision-making units which are party to the choice of a solution, the less likely the units will converge on a solution.
From a given starting point, the movements of an organization toward a solution may be described as a random, stochastic walk better than as linear progress.
The more problem attention, political interests, and policy communities converge on a solution, the more likely it is to be adopted as official policy.
- Corollary: This will be true even in comparison with other solutions which are more effective in cost-benefit or other rational-analytic terms.
Interest groups will spend more time and resources on blocking than on coupling activities.

Frequently Asked Questions

How is garbage can theory related to constructivist theory?
- March and Olsen (1989) discuss how non-rational "garbage can" policy politics can be rationalized on a post-hoc basis in a process of social construction of reality. Political institutionalization rests on the construction of common meanings emerging from irrational and incrementalist decision processes. In this way, "garbage can theory" is related to constructivist theories underlying institutional theory, discussed separately.
How is garbage can theory related to agenda-setting theory?
- John Kingdon, a developer of the multiple streams model, viewed the mass media as one of the multiple streams feeding in to agency policy-makers as they set their agendas. He explicitly incorporated elements of agenda setting theory and noted that media influence, on which agenda setting theory focuses, was correlated with public attention to issues. Multiple stream theory, however, gives much more attention to multiple determinants of policy agenda, including factors such as cost, interest group power, and the power of political parties and actors.

Bibliography

Bendor, J.; Moe, T, M; & Shotts, K. W. (2001). Recycling the garbage can: An assessment of the research program. American Political Science Review 95(1), 169-190.
Cohen, M. D., March, J. G., & Olsen, J. P. (1972). A garbage can model of organizational choice. Administrative Science Quarterly 17(1): 1-25.
Kingdon, J. W. (1984). Bridging research and policy: Agendas, alternatives, and public policies . NY: Harper Collins. Second edition: New York: Longman. Second ed., 1995.
March, J. G. & Olsen, J. P. (1989). Rediscovering institutions: The organizational basis of politics. NY: Free Press.
Melkers, Julia E.. & Willoughby, Katherine G. (1998). The state of the states: Performance-based budgeting requirements in 47 out of 50. Public Administration Review 58(1): 66-74.
Melkers, Julia E.. & Willoughby, Katherine G. (2001).Budgeters' views of state performance-budgeting systems: Distinctions across branches. Public Administration Review 6(1): 54-65.
Melkers, Julia E.. & Willoughby, Katherine G. (2005). Models of performance-measurement use in local governments: Understanding budgeting, communication, and lasting effects. Public Administration Review 65(2): 180-191.
Padgett, J. F. (1980) Managing garbage can hierarchies. Administrative Science Quarterly 25(4): 583-604.
Rubin, Irene (2006). the politics of public budgeting: Getting and spending, borrowing and balancing. Fifth ed. Washington, DC: CQ Press.
Rubin, Irene (2007). The great ujnraveling: Federal budgeting, 1998-2006. Public Administration Review 67(4): 608-618.
Sabatier, P. A. (1999). Theories of the policy process. Boulder, Colo.: Westview Press.
Thurmaier, Kurt M. & Willoughby, Katherine G. (2001). Policy and politics in state budgeting. Armonk, NY: N. E. Sharpe.