I can't think of a better public policy book than Nudge for the clear exposition of economic research, thoughtful suggestions, and downright humour. This is a book written for a country of diverse and strident opinions. Thaler and Sunstein suggest an incremental approach to public policy: why not nudge people into making better decisions for themselves. Leave them just enough choice to make them feel ownership over pension contributions, choice of medical insurance, even marriage. It's also a prescription for breaking the logjam in US Congress by taking a thoughtful, slightly right of centre approach to government. On my recent vacation to the West Coast I saw the ultimate bumper sticker for a cynical electorate: "Recycle Congress." ( )

Good read for those interested in UX, economics, and policy. ( )

http://pro-libertate.net/20090527/77-read-nudge-improving-decisions-about-health... ( )

The first few chapters are golden. The rest of the book is for those who cannot apply the first chapters on their own. ( )

The short version: Humans have cognitive biases that affect their decision making. Using what we know about these biases, we can design choice architectures that make it easy for people to make good choices without taking the freedom to choose away from those who want to do so. Thaler and Sunstein describe these principles and give examples of how they can be applied to saving money, health care, and preserving freedoms.

Nudge acknowledges both the failures of one-size-fits-all government solutions and why cognitive biases cause the market to fail as a mechanism for providing social good.

Details on the later since the former seems more obvious to me. The purpose of the market is to maximize profit. Ideally, this goal lines up with the goals of general social good. This is often the case because competition allows people to go with the solution that best meets their goals.

However, cognitive biases can throw a wrench in this system. Advertising is the most obvious way. The whole purpose of marketing departments is to figure out how to provide information in a way that takes advantage of cognitive biases.

Things still work out for the most part for choices that are frequent and give good feedback. No amount of marketing is going to make people like a bad candy bar. This is less true for decisions that are rare or have no feedback (e.g., health care, marriage, retirement investing, buying a house). Good decisions are hard because information is lacking, feedback is slow, and educating yourself can be difficult and confusing.

The authors suggest, and I agree, that a fruitful compromise is to use nudges. A choice architecture structured around nudges allows users to make choices but ensure that the easy decision is good enough for most users*.

An example: Many companies provide 401k plans with automatic enrollment. Getting out of these plans is easy enough, and this nudge has greatly increased 401k plan enrollment. Wonderful! The market is working all on its own**. But we can still make things better. When automatic enrollment is used, some default investment must be chosen. The default is often overly conservative because companies do not want to be liable for losses. Employees will often stay with this plan, even if they increase their contribution, because they assume the default choice is a good choice. A useful nudge here would be for the government to give best practices guidelines that suggest better investments and remove liability for losses for defaults which follow those guidelines.

That is just one example. The point is that by changing the choice architecture, we can create decisions that allow for choice while still making it easy to make a decision that is good enough. These nudges may come from the government or from the market, but both are necessary because they both have their own strengths and weaknesses.



* Yes, nudges could be used for bad as well as for good and you have to trust those making the choice architecture to make the nudged decision a good one without making alternatives too difficult. Given that, I still think nudges are better than one-size-fits-all solutions. I also believe that we already have nudges used for (sometimes) bad; it is called marketing. We may as well give the same tools to those who would use them to get rid of one-size-fits-all solutions.

** Not strictly true. There are laws that incentive companies to offer 401k plans, but we will ignore those for now. ( )