BEHAVIORAL BIASES AND HEURISTICS: A PRIMER Syon Bhanot, Swarthmore College World Bank DIME Workshop, Istanbul, May 12, 2015.

BEHAVIORAL BIASES AND HEURISTICS: A PRIMER

Syon Bhanot, Swarthmore College

World Bank DIME Workshop, Istanbul, May 12, 2015

Many Behavioral Challenges…

(Duflo, Kremer, and Robinson, 2011)

Many Behavioral Challenges…

(Duflo, Kremer, and Robinson, 2011)

Understanding Biases and Heuristics

Heuristics: simple rules used to make decisions “If it smells bad… don’t eat it” “If someone is running towards you at high speed,

yelling angrily… run away”

Heuristics can be either learned or hard-wired

Biases: when these mental shortcuts systematically lead us to make errors in our decisions

Nudges and Other Interventions

Behavioral interventions do not always work (or they can have very different effects for different types of people)

Context is key—which makes testing very important

Behavioral Concepts for Project Design

1) Status-Quo Bias and Defaults

2) Present Bias and Time-Inconsistency

3) Relativity and Context

4) The Limits of Choice

5) Social Norms and Motivators

1) Status Quo Bias and Defaults

This after an extensive education campaign and mass mailing to 12m of 15m residents of the Netherlands in 1998!

(Johnson and Goldstein, 2003)

1) Status Quo Bias and Defaults

This after an extensive education campaign and mass mailing to 12m of 15m residents of the Netherlands in 1998!

(Johnson and Goldstein, 2003)

1) Status Quo Bias and Defaults

Status Quo Bias – a preference for the current state

Can lead to: Inertia in decisions (financial, professional,

social, personal, etc.) Low take-up of important programs

Possible Solutions?

Defaults

Defaults in 401(k) Savings

(Madrian and Shea, 2001)

Active Decisions

Active Decisions in Financial Behavior

(Choi, Carroll, Laibson, Madrian,

and Metrick, 2009)

Smoothing the Path of Least Resistance

Home delivery found to raise fertilizer use by 70%.(Duflo, Kremer, and Robinson, 2011)

Vaccination and implementation intention prompts(Milkman, et. al., 2011)

Now, A Question…

Would you rather have a professional, 20-minute massage right now or a

professional, 25-minute massage tomorrow?

Most people want it now…

Now, Another Question…

Would you rather have a professional, 20-minute massage on May 12, 2016 or a

professional, 25-minute on May 13, 2016?

Now most people are willing to wait an extra day…

2) Present-Bias and Time Inconsistency

Humans struggle with self-control, overweighting the present in their decisions

Many reforms involve long-run benefits and short-run costs or hassle

So present-bias can prevent program take-up or deter otherwise beneficial actions

Possible solutions?

Commitments to Bind Us

Commitment Savings in the Philippines

(Ashraf, et. al., 2006)

Save More Tomorrow(Thaler and Benartzi, 2004)

Innovative Mobile Platforms(Digit, Acorns, etc.)

Using Our “Future Selves”

Encourage Long-Term Thinking

(Hershfield et. al., 2011)

3) Relativity and Context

Nearly every decision and choice is dictated by how one thing appears in relation to another

Poor Doug Pitt…

Rome Paris

Vs.

Asymmetric Dominance

(Ariely and Wallsten, 1995; Huber, Payne, and Puto, 1982)

Free Coffee Included!

Rom

e U

tilit

y

Paris Utility

R+

P+

Asymmetric Dominance

R

Decoy

Rom

e U

tilit

y

Paris Utility

R+

P+

Asymmetric Dominance

R

Rom

e U

tilit

y

Paris Utility

R+

P+

Asymmetric Dominance

PDecoy

Rom

e U

tilit

y

Paris Utility

R+

P+

Asymmetric Dominance

P

Att

rib

ute

1(c

ost)

Attribute 2(caffeination) (Simonson, 1989)

The Compromise Effect

Att

rib

ute

1(c

ost)

Attribute 2(caffeination) (Simonson, 1989)

The Compromise Effect

4) The Limits of Choice

More choice is not always good – “Choice Overload”

4) The Limits of Choice

Having more fund options in individual 401(k) savings plan is associated with less investment in equity (Iyengar and Kamenica, 2010)

In consumer behavior, more choice can mean more curiosity… but less actual purchases (Iyengar, 2000)

However, when looking at Medicare plan choice, evidence that more choice leads to more beneficial switching (Ketcham, et. al., 2015)

5) Social Norms and Motivators

We care about how we compare to those around us

This can be a powerful tool to influence behavior at low-cost

5) Social Norms and Motivators

Allcott (2011): Finds social information mailers reduce energy use by ~2%

Kraft-Todd, et. al. (2015): Review of literature finds social motivators a promising tool to promote cooperative behavior in real-world settings

5) Social Norms and Motivators

Some limits and heterogeneities in response to social information: Beshears, et. al. (2014): Peer savings

information decreases savings for previously low-savers

Schultz, et. al. (2007): Evidence of “boomerang effects” from social information

Bhanot (working paper, 2015): Competitive framing on social information can demotivate low-performers

Key Takeaways

1) Human behavior and decisions are at the core of all projects being discussed here this week, in some form

2) Humans suffer from behavioral biases that keep them from making logical/rational decisions in some contexts

3) By understanding these biases (and testing them using impact evaluation methods), we can improve project design, increase take-up, and maximize impact

Thanks!

Syon BhanotSwarthmore College

Contact:

spbhanot@gmail.com

scholar.harvard.edu/sbhanot