Skip to content

Lecture 17 A few effects & Choice Architecture

Sairam

  • Addiction and Abstention
    • Conventional model: More consumption \(\implies\) more total utility
    • NM talks about 'bads'. Increasing consumption is monotonically bad.
    • For some people, \(\exists\) goods that give too much pleasure, though excessive consumption \(\implies\) severe issues
    • Addition is a major factor (alcohol, tobacco, recreational drugs)
      • Also, junk food, gambling, computer games (any activity with significant degree of excitement)
    • Psychological and Physiological mechanism related to this enjoyment is complex. Double-edged nature
    • Studies show that people infer happiness from their actions in a self-signaling manner. \(\implies\) Diagnostic Utility
      • e.g. Daily jog despite a rain \(\implies\) Gratifying signal of willpower, dedication or future well-being
    • To indulge may reveal themselves to have an addictive personality. Acknowledgement of this weakness \(\implies\) impossible to break the addiction
    • "Falling off the wagon" can be explained by DU and self-signaling \(\implies\) Low self-esteem and indulgence

Let's discuss the reason for the inertia

  • Endowment effect
    • Utility depends on possession also
    • People acquiring a good through purchase or as a gift, tend to value it more highly than others (who have not acquired it)
    • Read Endowment effect and Loss aversion
  • Status Quo Bias
    • General tendency to stick to one's current situation
    • Less cognitive effort to repeat the previous decision. (e.g. Restaurant default order, customer favorite)
      • A confound1: Estimation \(P(\text{Prev Order Good})\) is higher than selecting a previously untried option.
    • E.g.
      • seating preferences in class
      • 7-day free trial or subscription
      • scheduling shows in TV channels
    • Happens due to lack of attention
    • \(\implies\) important policy implications, sales strategy and government intervention.
  • Introduction to Choice Architecture
    • SE Policies rely on 'hard' measures2.
    • Hard measures are desirable when behavior is detrimental and should be avoided
    • Alternatives: soft measures, through choice architecture. Described by careful design of how choices are presented.
  • What is Choice Architecture?
    • Choice architect: indirectly influence others' choices
    • To ensure: the Automatic system doesn't get confused
    • Choice Architecture:
      • Strategic arrangement of DM environments to influence choice without restricting freedom
      • without altering the underlying objective of the choice
      • Carefully structuring the context to gently guide them to desired outcome

Principles of good choice architecture

Say, we are the choice architect

  • Defaults - The Path of Least Resistance
    • FE
    • We may use availability heurist 'in the framing' to nudge.
    • 'defaults' is a popular framing that exploits 'status quo bias'
    • Most people will take whichever option requires least effort
    • If \(\exists\) default option (that gets selected on doing nothing), the most people will end up with that option
    • 2012: UK, under 2008 Pensions Act, firms automatically enroll employees in a pension scheme.
      • If employees wanted to opt out, they are free to do so. But, since it requires effort, cognitive resources, time and even psychological conflict, they leave it at that.
    • \(\implies\) Cognitive laziness so not many people will opt out.
    • e.g. "Regular" or "custom" installation (one of the boxes is already checked, indicating default)
    • Defaults are unavoidable!
      • For any CAS, \(\exists\) associated rule... "What happens if the DMaker doesn't do anything"
      • If choice is complex and difficult, people appreciate a sensible default (e.g. Vim)
    • For a consumer, a default may be costly. e.g. "Our most popular plan". If we had researched for just a few minutes, we would have not wasted ₹2,400.
    • In other cases, may be beneficial: opportunity cost of our time
  • Expect Error - Design for Human Mistakes
    • A well-designed system is as forgiving (to humans erring) as possible
    • ATMs ask uses to remove their card before spitting out the money. "Please remove your card"... a strategy called forcing function ("In order to get what you want, do something else first")
    • Cars come with warning systems: passenger safety operations related to headlight / when about to run out of fuel / oil change needed
    • For elderly people, design meds as once a day. Educate the Automated System to think: "My pill(s) every morning, when I wake up". This forms a habit
    • US Visitors to London. Opposite road-side rules. They are helped by good design: Pavement signs: "Look right!"
  • Feedback - Helping People Learn and Adjust
    • Tell when they are doing well, and when they are erring
    • Modern cameras: take care of not removing lens cap, not placing target object at the centre.
      • Includes, very satisfying but fake "shutter click sound"
    • Ceiling Paint: pink when wet, white when dry.
    • Laptops: Plug in, battery low.
    • Problem: You supply so many warnings that they get ignored.
  • Mapping - Translating Choice to experience
    • Mapping is the relation: \(\text{Choice} \to \text{Welfare}\)
    • Good system, help them map and select options that make them better off
    • Hard to track how much credit card usage costs (complex pricing schemes)
    • Libertarian paternalism gives a mild form of government regulation. RECAP: Record, Evaluate, Compare Alternative Prices
      • Requires firms to disclose all fees in a standardize, machine-readable format
      • Help consumers easily review actual usage and compare pricing plans without restricting choice
  • Structuring Complex Choices
    • As alternatives become more numerous, complex, we have to think more and work more as we are more likely to influence people's choices, for better or worse
    • Strategies that can be used:
      • Elimination by aspect (filtering?) #doubt
      • Organization by paint shops (color wheel)
      • Collaborative Filtering: Use judgements of other people who share your taste to filter. E.g. books or movies, to increase the likelihood of picking the one you like.
        • Wisdom: it's good to learn what people unlike us like, to see whether we might like that or not.
    • Structuring means helping people to learn so they make better choice on their own
  • Incentives
    • Sensible us, will put the right incentives on the right people
    • Ask four questions
      • Who uses?
      • Who chooses?
      • Who pays?
      • Who profits?
    • Free market incentivizes, good products' production and sell them to the right people due to competition
    • Incentive conflicts:
      • e.g. if I pay for the entire group lunch, vs someone else paying... I will go for more expensive food.
      • Exist in many markets including healthcare
        • Patient uses (Just wants to get treated, only incentive that should matter)
        • Physician chooses (different incentive)
        • Insurance company pays (different incentive)
        • Equipment manufacturers, drug companies, malpractice lawyers get the profits (different incentive)
    • Salience (prominence)
      • Most important modification that must be made to standard analysis of incentives.
      • Grab the attention to incentives in their choices
      • e.g. To control people's behavior, instead of just quietly raising the electricity price, let the AC notify them about the cost per hour of lowering the temperature a few degrees during a heat wave.

  1. An alternate reason that may exist, but its effect is difficult to isolate 

  2. Regulations, bans, penalties, subsidizing