In Multiple Streams Approach, the observation that policy-makers do not act ‘rationally’ (why? see bounded rationality) but instead rummage around in the ‘garbage can’ into which old unused/failed/outdated/novel policy solutions have been tossed, looking for a (quick) fix for a problem that has cropped up (also, sometimes people look for problems to which their pet solution can be applied).
The garbage can is “is a collection of choices looking for problems, issues and feelings looking for decision situations in which they might be aired, solutions looking for issues to which they might be the answer, and decision makers looking for work”
From a famous 1972 paper by Cohen, March and Olsen
And here’s a video in which one of the authors explains its origins
See also http://faculty.babson.edu/krollag/org_site/encyclop/garbage_can.html
See also tail wags dog and the ‘if the only tool you have is a hammer, all your problems begin to look like nails’ thing…
See also Primeval Soup (MSA), Policy Stream (MSA)
Although the garbage-can model captures some aspects of real life, it puts too much emphasis on randomness. Participants normally perceive actions as being solutions to problems, and they perceive logical sequences of activity that produce these problem-action associates. The problem-action associates and activity sequences appear non-random to the participants, and a theory of organizational action should explain why participants perceive certain activity sequences and certain problem-action associates. The garbage-can model also overstates the importance of problems by treating them as being just as important as actions. Actions are what organizations are all about, what organizations are designed to produce. Organizations could exist happily without problems if it were not that societies insist that organizational actions ought to solve problems.
(Starbuck, 1982:16)
Starbuck, W. 1982. Congealing Oil: Inventing Ideologies to justify acting ideologies out. Journal of Management Studies, Vol. 19, (1), pp. 3-27.