Revisiting the Logic of Consequences 177
stereotypes and typologies that simplify what information is and is not consid-
ered. Problems tend to be unbundled and reduced to their parts. It is oft en easier
to associate particular available information with the parts of a problem in the
search for more comprehensive solutions; this unbundling is especially useful in
the context of organizational specialization and departmentalization. Given avail-
able information, parts of problems can be managed in this disaggregative form.
Some information, such as budgets, balance sheets, and performance mea-
sures, is always given a special place in decisionmaking. Th is information has
the authority of evident objectivity and certainty. Decisionmakers give careful
guidance to organizing and developing this type of information source and thus
manage the information upon which their future decisions are to be based.
Coping with imperfect information is bounded rationality and satisfi cing in
action. Satisfi cing, or “good enough” rationality, has the very big advantage of
moving an organization in the direction of preferred values while preserving in-
stitutional equilibrium. Indeed, recent work on the application of bounded ratio-
nality to decisions by street-level bureaucrats indicates decisionmaking is further
shaped by information about what others within the organization will do (Keiser
2010). Bounded rationality, thus understood, is rational behavior that stabilizes
and supports continuity and order while enabling at least some adaptation. Policy
scholars have shown how such a framework applies to the act of public budgeting
(see Davis, Dempster, and Wildavsky 1966; Jones et al. 2009).
Even with perfect information, however, decisions will depart from pure ra-
tionality. Empirical evidence demonstrates that information processing is also
shaped by cognitive biases and the behaviors of others. As noted earlier in the
chapter, evidence from behavioral economics shows a strong tendency toward
a status quo bias (Tversky and Kahneman 1974). Past information and previ-
ous decisions exert a powerful eff ect on future decisions. Although such evidence
may appear at face value to be similar to incrementalism or bounded rationality,
the underlying assumptions are diff erent. Even in cases of complete information,
the processing of information will in large part be shaped by previous decisions.
Th is tendency toward relativity bias or an anchoring eff ect (Ariely 2009) shows
that future decisions and, importantly, the evaluation of alternatives tend to be
based on previous decisions, experiences, or outcomes. Bounded rationality is
oft en linked with incrementalism, in which incomplete information prevents full
consideration of alternatives, and thus the safe play is to make small, incremental
adjustments (Jones and Baumgartner 2005a, 326). Cognitive biases, such as an-
choring eff ects or a “commitment bias” (Brafman and Brafman 2008, 26–39), can
lead to decisions that result in no change in policy despite new information.
Attention
Attention, both individually and collectively, is a scarce resource, refl ected in
limitations of time, too much information, changing problems, and changing