March, James G. A Primer on Decision Making, p. 95-100

Chapter 5:

Competency Traps

The more a particular rule is used, the more likely the rule will be reinforced by experience. The more its reinforc ed, the more it will be used. Thie positive feedback loop reduces the possibility of substituting another, potentially more suitable alternative.

This can result in technological "lock-ins" at the individual, organizational, and societal level.

Interactive Effects on Outcomes

Interactions among learners further complicate things. The outcomes for any particular action depend on what other decision makers do. Decision by any one person often have to take into account the preferences, identities, and likely actions of others. Rules are affected by other rules.

In competition, the outcomes of any particular competitor is a function of its alternative actitivities and those of other competitors. Behavior is strongly influenced by the ecological structure. Some decision makers may affect others without themselves being affected. They can act as predators and prey. Or they could cooperate. The frequency one uses a particular technology may largely depend on what others are doing.

These network externalities can make autnomous learning misleading.

Aspirations, Diffusion, Legitimacy

Aspirations are social. The evaluation of a given performance depends on the performance of others. It increases the chance that outcomes will deviate from aspiration (vs self-referential aspirations) and decreases the likelihood of changing from success to failure and vice versa. The population may diverge into two groups (successes and failures), and learning may become superstitious.

Rules also reproduce by diffusion through social networks via knowledge transfer. Networks include associations, consultants, employees transferred, educational institutions, etc.

The legitimacy of rules is also affected by the use of rules by others. Appropriate behavior is socially constructed. People observe others to discern and legitimize rules. Commonly used practices become myths. Official bodies may also legitimize rules, as may professional organizaitons defining SOP's . One way a rule becomes legitimized is its use in another.

James G. March, A Primer on Decision Making, pg 221-272.

Chapter 6 Decision Engineering

Defining Decision Intelligence

Often past decisions are viewed as rational if the outcome was good. Or they are seen as procedurally rational if the process chooses actions that are expected on average to lead to desired outcomes. Organizaitons are seen to have learned if their performance improves.

Convictions that good outcomes will be achieved if decision makers follow preferred processes underlie much of the perscriptive literature.

But the intelligence of an action is defined in terms of outcomes. It is intelligent if it satisfies whe wishes of relevant parties. It makes it ex post and subjective, sometimes making evaluation of intelligence difficult.

Tradeoffs of Intelligence

Symbols and Substance

Decision making allocates scarce resources and displays values and interprets life. They exhibt and reassert social beliefs. Gathering information and making decisions are signals and symbols of competence. Decision makers gather information and do not use it; ask for more and ignore it; make decisions first and look for relevant information afterward, and gather and process information that has little or no direct relevance to decisions. Sometimes it's best to buffer the symbolism fromthe decision making. Substantive choice are often made outside formal decision processes. Or decision makers can improve the symbolism and maintain the best social beliefs and myths about choice, reason, power, conflict, and intelligence.

Comparisons Across Nested Time

Actions good for the short run are often bad for the long run. Preference and identities change over time, partly as a result of taking actions. Should outcomes be evaluated with preferences that existed at decision time or at evaluation time? Delayed gratification is difficult.

Should intelligence be based on values at outcome time or decision time? Decision makers may also need to consider changes to values with action as well as potential outcomes in making their decision.

Comparisons Across Nested Decision Makers

Intelligence is difficult to measure when it involves multiple assessments. Interpersonal comparisons of value create unresolvable problems. How do you weight the opinions of others? But the system of limited attentions seems to help in these cases -- coalitions, bargaining, exchange also help. Deviding into departments separates actions and outcomes from full scrutiny by all interested parties. Inconsistencies are ignored through ignorance. Limited attention sacrifices global consistency for local consistency.

Also decisions favorable to groups may not be favorable to the system as a whole, and vice versa.

Decision Making Myopia

Out of Sight, Out of Mind

Decision makers are usually less attentive to distant consequences. But sometimes favoring the short-term is necessary for survival.

Apostles of Reckoning, Apostles of Progress

The reckoning folks see the world's resources as fixed. Gains at one time will be paid for at another time. The progress folks see the world's resources as expandable. Divine judgment means rewards on earth are met with penalities in heavan. Economic exploitation sees rich in on place paid for by poverty elsewhere.

On the other hand, religious conversion sees that each new person led to the Truth adds total virtue to the world. Scientific progress increases over time and leads to increases in assets and resources. Gains of one can lead to gains for all.

Improving Adaptiveness

Decision procedures are rarely perfectly adaptive to environments. Some inefficiencies are:

* lags in matching to enviroment

* multiple equilibria (local maxima)

* path dependency

* networks of diffusion (outcomes depend on way information spreads)

* mutual adaptation (decision makers and environment adapt at same time)

* ecologies of adaptation (coevolution)

History is a locally adaptive, branching process with multiple equlibria.

Extracting knowledge from experience

Individuals have limitations in interpreting evidence. They learn lessons inadequately, recall memories incompletely, estimate futures inaccurately. They learn superstitiously. They face four structural problems:

1. Paucity of Information

Ameliorated by:

* rich histories

* multiple observers and multiple interpretations

* hypothetical histories

2. Redundancy of Information

Past actions tend to be repeated . leading to observations based on the same or similar experiences. They repeat successes. Ways to reduce redundancy are:

* increase measurement noise (leading to arbitrary failures)

* reduce retrival of memory

* inrease rate of aspirations adjusting

3. Ambiguity of Information

It's hard to isolate causal effects from random or extraneous forces. Reducing ambiquity is done by:

* making big changes rather than small ones

* slow the rate of adaptation to experience

Extracting Knowledge from Others

Imitation can spread both good and bad ideas. But bad ideas spread more rapidly among the ignorant than the informed, and good ideas spread more quickly among the informed than the ignorant. But firms need knowledge inventories and capabilities to appropriate and exploit the information of others.

Profiting from the Rules of others

Following rules or forms can gain legitimacy.

Matching situations to repertoire inventories

Often people don't recognize a problem until they find a solution. They look for match between a situation and practices, preferences, and identities they have in their repertoire library. Monitoring systems thrive on informal networks, gossip, etc to get early warnings.

Complications of Knowledge

Knowledge Substitutes

Substituting Technique for Creativity

Experience tends to protect beliefs and strengthen existing paradigms. Impovements of technique within a schema tend to confuse comparisons between schema.

Substituting Power for Learing

Dilemmas of Competitive Disadvantage (draws from distributions ability vs luck)

Politics of Knowledge

Strategic Nature of Knowledge

Systems of information are instruments of power. People try to assure that the values and preferences of the knowledge providers are similar to the decision makers.

Nonneutrality of Knowledge

Creating Meaning

Expanding Awareness