Notes for "Nudge" by Richard H. Thaler and Cass R. Sunstein

February 19, 2017


My introductory example is The introductory example is placement of food in school cafeterias. The take away is people can be influenced by changes in context and this power, the power to influence can be used to benefit those with little cost to those benefitting from it.

Your decide which is better.


I recommend to understand the terms used first, but its only [just a suggestion]({{< ref "/thoughts/" >}}).

Read these if you want a feel for the terminology now or to refer back as needed to clear up any confusions around what these words may mean.

Nudge : Physical definition: To push mildly or poke gently in the ribs. Alternative definition more describe the book's usage: One who nudges is to "to alert, remind, or mildly warn another". : For this context: preferences between choices where the cost between said choices is low. Not a mandate. Ex: the example of eye level fruits vs outright banning of junk food.

Libertarian Paternalism : Libertarian in that people should be free to choose. Such as policies which either maintain choose or expand the freedom of choice. — "Liberty-Preserving" : Paternalistic in that self-conscious thought or design goes into directing people to choices which can be judged by themselves as markedly better.

Choice Architects : Have a responsibility for organizing the context in which people make decisions and must adapt to requirements of the context — "There is no such thing as a 'neutral design'".

Econ's : The naive descriptive view of people being perfectly rational decision makers and can make unbiased forecasts of their environment and choices.

Humans : Are the more accurate model of peoples behavior, in that we succumb to many fallacies, rely on status quo biases and, have finite resources to dedicate to making decisions.

The Ideas

Default Choices

Never underestimate the power of inertia

People have a strong tendency to go with the status-quo or initial default options. Power lies in those with the ability to make well-chosen defaults.

Misguided Opponents to Paternalism

Opponents fall for the misguided approach of trying to maximize the statistic of total number choices, and anything less might as well be the dreaded overly handed single mandate.

This is a local optimum well chosen from their perspective of time involved. I.e. why should I think about how to organize choice, when I'm not the eventually decider.

But is easily refuted via simply imagining a game or sport situation where a novice is up against the pro. Predictable is the outcome, in that the novice eventually loses due to inferior choices.

Novices vs "The Pros"

Our day to day choices as people can be modeled as we, being the consumers, interacting and going up against experienced professionals, "The pros," who are selling us many a trinkets and wares. Most of us have deep but narrow expertise in a small number of domains out of the large number of varying contexts we interact with.
A reasonable assumption is that we can make good choices in domains we have experience in, but do far less well in contexts in which we are poorly informed, lack experience, or the feedback given to us is slow or infrequent.

Misconceptions of Paternalism

  1. That it is possible to avoid influencing people's choices.
  2. That paternalism always involves coercion. Valid points are that governments cannot and should not be trusted to actually be competent or benign. A real fear can that elected officials will place their own interests first, or pay attention to a narrow set of goals from private groups. Shared in these concerns are the authors as they agree that risks of mistakes, and overreaching of government are real. For this reason they favor nudges over the inferior alternatives of: commands, requirements and prohibitions. Insisting that choices remain unrestricted is a good regularizer to keep the risks of inept or corrupt designs reduced.

    Freedom to choose is the best safeguard against bad choice architecture.

Choice Architecture

Basic example is knowing that humans are choosers with limited attention and resources to make decisions. Therefore the designs of a choice architecture do the upfront work of making decisions. Designs which minimize the costs imposed others.

Humans & Biases

We make distinctions between two types of thinking, one intuitive and automatic, and the another reflective and rational.

Biases are important to understand due to the limited time and cognition people can apply to choices in their day to day lives. We often make doe with and accept questions posed as is rather than trying to determine whether their answers would vary under alternative formulations.

System 1 or the Automatic System

  • Rapid and feels instinctive.
  • Not involving deliberation or mental resources.
  • Associated with oldest parts of the brain.
  • Can be trained through massive amounts of repetition.

Being truly bilingual means that you speak two language using the Automatic System.

One reason why teenagers are such risky drivers is that their Automatic Systems have not had much practice, and they must rely on the slower Automatic System.

Heuristics or "Rules of Thumb"

Anchoring, Availability and representativeness


Taking what number I know and adjusting the direction from that number till it fits the problem or question at hand. Obviously biasing towards the initial number.

Anchors service as nudges, via a subtly suggestive starting point to shape my initial thinking


To answer questions requires looking up or assessing likelihoods of risk and theorizing and then searching for readily example. This heuristic is often used when calculating risk. Where an individual would weight recent events as more probable.

Our Automatic system is too quick to stop checking farther than the recent past and doesn't automatically continue its search without explicit motivation from our Reflective System.

The availability heuristic helps to explain much risk-related behavior in both public and private decisions to take precautions.

Misperceptions can affect policy, because governments are likely to miss allocate resources if they respond to people's fears rather than the more probable dangers.


This is a similarity heuristic. Easier said with an example, so imagine being asked to judge two things, Thing An or Thing B. The question is wether or not Thing A belongs in the same category as Thing B. What the heuristic describes is how we use our Automatic Systems to quickly see how well A matches with the stereotype of B. So really we are asking ourselves how representative Thing A is of Thing B.

This line of thinking isn't wrong per se, but is a bias when the similarity and actual frequency of Thing A being in category B diverge. So Thing B and Thing A might often be in the same category, but when taking in consideration all categories, Thing A has higher probability of being in another.

We often see pattens because we construct of informal tests only after looking at the evidence. Basically saying find comfort in correlating what we can easily see or measure, but not what actually produced said evidence. The example given is the WWII Londoners seeing patterns in the bombing runs, and thinking the germans were deliberately avoid their spies, when in actuality it was purely random. Basically they were confusing random fluctuations with casual patterns.

Optimism and Overconfidence

Everyone looks to take the pre-class survey and say they will be in the top 10% at the end. They have a pervasive, "I'm going to be above average".

People are unrealistically optimistic even when the stakes are high. Unrealistic optimism characterizes most people in most social categories.

Gains and Losses

People hat losses, and our Automatic system will make us very emotional about them. Loss aversion operates as a kind of cognitive nudge, pressing us not to make changes, even when changes are very much in our best interests. It's a form of inertia.

Status Quo Bias

Another form of inertia comes from our general tendency to stick with our current situation. Easily demonstrated by the tendency to students to stick to a similar spot even with unassigned seats.

The combination of loss aversion with mindless designs after any initial deliberation, implies that if an option is designated as the default it will attract a large market share.

Default options thus as a powerful nudges.


EX. 90 of 100 are alive vs 10 of 100 are dead.

Framing works because people tend to be somewhat passive decision makers. Their Reflective System does not do the work required to check and see if reframing the questions would produce or inform of another answer.

Resisting Temptation

Example given is dinner guests mindlessly eating through a bowl of cashews. The group began eating wanting the nuts, but we equally relieved when it was pulled away, so as to not spoil their appetite.

This example demonstrated a group behaving dynamically inconsistent . Meaning that initially the preferred option A over B, but after some time, they then prefer B over A.

A simple solution is given by Ulysses and his tactic to overcome the Sirens.

Most of use lack good self control primarily due to use underestimating the effect of arousal. In this sense its arousal is being similar to “being in the moment”. This is described as the hot-cold empathy gap,

  • Hot describing the aroused state, the automatic system.
  • Cold describing the calculated planner, the reflective system.

What ends up happening is some parts of the brain are tempted and other parts are prepared to resist temptation. This can causes server inner conflict.

Mindless Choosing

This is shown via the Cashew, stale popcorn, Sunday morning driving to the office, and tomato soup examples. All demonstrating how in many situation we put ourselves into automatic pilot mode and we are not actively paying attention to the task at hand.

Self-Control Strategies

Our planners taking steps to control the future actions of our Doers, often by changing the incentives or minimizing the damage from being in automatic mode. Examples are the alarm clock which runs away, loss aversion for the Phd student, and weight loss bet between friends.

Mental Accounting

Is the system (implicit or explicit) where we evaluate, regulate, and process our resources in a way to control usage of that access. Resource in this sense is usually money but could be anything where the mental accounting scheme takes something usually fungible, meaning without labels, and constrains it, operating it out into different groups. Thereby making it non fungible and limiting it’s allocation and usage. Aka the rainy day jar, the Christmas savings account, etc.

What are the dimensions of our choice environments?

Benifits, Costs, and when

When our choices create immediate costs, with an eventualy but delayed, benifit its hard to make appropriate descisions.

Difficulties in complexity

Easy things are automatic, while the hard things require managing many thoughts at once. Some further reasons for difficulty include

Frequency, a lack thereof

Many big life altering decisions are rare, and are done without much practice.

Feedback and learning structures

Long term processes rarelly give good feedback.

Gauging eventual experience

Interpretation of what the outcome will be after a choice is made is often not knowable. Examples; foreign dinner menu, Entertainment when watching an new sport.

Matthew Clemens © 2022