For anyone who wants it, there is lots of advice around the place on how to improve decision making. There are hundreds of books on the topic, innumerable articles and magazine pieces, and unstoppable streams of ephemera on the internet.
One thing I’ve noticed about this flood of advice is that each theorist’s thinking tends to be dominated by a particular “take” on what decisions are, what decision making is like, and how it should be done. But the field of decision making is a bit like the Melbourne Zoo – there is not one type of animal but rather hundreds of different types, most falling into one broad category or another. Nowhere at the zoo do you see a sign pointing to Animals; only directions to the reptiles, the birds, the butterflies, etc.. Likewise there is not much you can say about Decisions In General; anything interesting or useful needs to focus on a particular type of decision.
Such as? What kinds are there?
Given the rich heterogeneity in the decision arena, there are many ways to “slice and dice” the domain. How you do so should depend on what you want to achieve, where, and why.
Lately I’ve been doing work with some large organisations looking at their decision making and how to improve it. In this context, it seems useful to distinguish at least the following four major kinds, based primarily on the type of activity (cognitive, and/or social) that goes into making the decision.
1. Intuitive Decisions
As elsewhere, the great majority of decisions in large organisations are intuitive. Intuitive decisions have the following characteristics:
- They tend to be made rapidly
- They are made by individuals deciding alone
- There is little or no conscious reflection
- There is little or no following of rules or procedures
- Usually only one option is considered
- The cognitive process is typically “recognize/act” – the decision maker recognizes the situation and immediately “knows” what an appropriate thing to do in that situation would be
2. Technical Decisions
Technical decisions are those made by following some well-defined technical procedure. Very often a spreadsheet or computer program plays a key role in the decision processing, and indeed sometimes the decision making can be largely or completely turned over to a computer. Examples include:
- Using multi-attribute utility theory to choose which IT system to adopt
- Using formal decision analysis to select a strategy for a negotiation
- Deciding whether to make a loan, or how much to loan, based on the applicant’s information as provided on a standard loan application
3. Deliberative Decisions
Deliberative decisions are those made by “weighing up” the arguments and evidence (“pros and cons”). Deliberative decision making:
- Involves conscious articulation and evaluation of various options and the considerations counting for and against options
- Takes anywhere from minutes to months
- May be made individually, or by a group
- Often involves discussion and debate
- Doesn’t involve computers in any important way, since the relevant considerations are usually “informal” or qualitative cannot be rigorously specified or quantified
- Is not governed by strict rules or procedures, though it is subject to various norms and conventions (usually unspoken or implicit)
3. Bureaucratic Decisions
These are decisions which an organisation makes using a “bureaucratic” procedure. Typically they:
- Are important/weighty/high stakes
- Are “standard” sorts of decisions for the organisation; hence they
- Are made following a codified process, involving lots of steps and rules
- Involve lots of people playing their assigned roles
- Are based on lots of information, extensively analysed
- Have detailed documentation
- Making major decisions in military operations using the Military Appreciation Process (or similar)
- Making a major infrastructure decision following Infrastructure Australia’s Reform and Investment Framework
- Making a substantial investment decision in a funds management firm, following that firm’s standard procedures
(Note: I don’t claim that this list of decision types is “MECE“. Indeed it is probably a more useful classification than any strictly MECE list would be.)
Once these various types of decisions are distinguished, you can see that asking a question like “How can decision making be improved?” is bit like asking “What do we feed the animals in the zoo?” It all depends what type of decision you’re interested in. At the most general level, you’d do very different things to improve decision making of the four types described above:
- Intuitive decisions could be improved through more and better experience, feedback, possibly using some debiasing techniques
- Technical decision making could be improved through better selection of technical methods, and training in those methods
- Deliberative decision making could be improved by adopting more effective deliberative practices and supporting technologies; and
- Bureaucratic decision making could be improved by revising the official procedures.
Five ways regulators ‘think wrong’ in the WSJ.
Good article Tim. I also think though that Some classes if bureaucratic decisions (eg MAP) could also be improved by technology. Thanks for your thoughts. -Simon Pope
Tim. good work what a help you are!