brigitte madrian harvard university

BEHAVIORAL ECONOMICS AND

FINANCIAL REGULATION

Brigitte Madrian

Harvard University

December 7, 2017

Behavioral Economics and Public Policy 2

What is Behavioral Economics?

Behavioral economics incorporates insights from economics and other behaviorally oriented disciplines including psychology, sociology, anthropology, and cognitive neuroscience, to enrich standard economic models in ways that improve our ability to understand and predict human behavior, market outcomes, and public policy.

3

What is Behavioral Economics?

Traditional Economics Behavioral Economics

4

Neurology of Decision Making:

Multiple Systems Hypothesis

The brain integrates signals from multiple systems in making decisions

These systems process information differently

System 1 System 2

Affective Analytic

Fast Slow

Reflexive Reflective

Effortless Effortful

Impulsive Deliberative

Myopic Patient

5

INFORMATION

FAILURES

Traditional Economic Motivations for

Policy Intervention 6

MARKET

POWER



EXTERNALITIES



PUBLIC

GOODS



Behavioral Economics and Public Policy

What does an understanding of

behavioral economics bring to the policy table?

Additional motives for policy intervention

New policy tools

Ways in increase the effectiveness of

traditional policy tools

10

Behavioral Economics and Household Financial

Outcomes 11

Where Do Traditional Economic Models

Fall Short?

Limited attention

Selective attention

Salience

Limited computational capacity

Choice overload

Mental accounting

Use of heuristics

Biased reasoning

Errors in probabilistic

thinking

Motivational biases

People are Imperfect Optimizers:

They Make Mistakes!

12

Salience: Consumers Pay More

Attention…

…to up front fees …than to shrouded fees

Air travel ticket price Air travel add-on fees

Product prices Sales tax or shipping costs

Bank interest rates Other bank fees

Cell phone monthly fee Cell phone overage charges

Mental Accounting

241

414

0

100

200

300

400

500

One envelope Two envelopes

Rup

ees

saved o

ver

14

weeks

One vs. Two Envelopes

Source: Soman and Cheema (2011). “Earmarking and Partitioning”

14

How Much to Save: Economics…

15

0

1

2

3

1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15%

Frac

tio

n o

f e

mp

loye

es

Year-End Contribution Rate

Source: Choi, Laibson, Madrian and Metrick (2006)

…vs. Psychology

Heuristic thinking: people disproportionately choose to

Save at contribution rates that are multiples of 5

16

Future Value of $1000 Invested Today

0

2.500

5.000

7.500

10.000

12.500

15.000

17.500

20.000

0 5 10 15 20 25 30 35

At a 10% interest rate

compounded annually

At a 10% interest rate

with no compounding

Heuristic thinking: people anchor on linear

compounding, and adjust upward, but not completely

17

Procrastination

Overconfidence

Addiction

Affect

People Have Self Control Problems:

They Cannot Always Follow Through with Their Plans

Where Do Traditional Economic Models

Fall Short? 18

Mood and Tipping

19% 24% 27% 29%

Source: Rind (1996). “Effect of Beliefs About Weather Conditions on Tipping”

Average tip given to hotel room service delivery person

by the weather conditions reported to hotel guests.

19

Where Do Traditional Economics

Models Fall Short?

Reference-dependent preferences

Endowment effect

Loss aversion

Status quo bias

Menu effect

Framing

Other-regarding preferences

Altruism

Fairness

Social norms

Interpersonal preferences

Non-standard preferences: what individuals want is not

what traditional economic models presume they want

20

Loss Aversion and Investor Decisions

Individuals treat

investments that have gone

up in value differently

from investments that have

gone down in value

More likely to hold onto

stocks that have fallen in

value

Set a higher list price

(riskier strategy) for real

estate

21

Default/Endowment Effects and Asset

Allocation Outcomes

23% 20%

95%

27%

0%

25%

50%

75%

100%

Match default: employer stock

Match default: none

Alloca

tion

to E

mplo

yer

Sto

ck

Own

Match

Employer Stock Allocation in Employee Savings Accounts

IPO Lotteries in India

Source: Choi, Laibson and Madrian (2009)

63,9%

45,8%

36,6%

1,2% 1,7% 1,5% 0%

20%

40%

60%

80%

At IPO 12 months later

24 months later

Hold

s IP

O S

tock

Won IPO lottery

Lost IPO lottery

Source: Anagol, Balasubramanian and Ramadorai (2016)

22

Implications for Financial Market Regulation 23

The Interaction between Individual

Behavioral Biases and Firm Motives

Behavioral

Fallability

Market neutral and/or

wants to overcome

consumer fallability

Market exploits

consumer fallability

Consumers

misunderstand

compounding

Savings context:

• Banks would like to help

reduce this bias

Borrowing context:

• Banks would like to

exploit this bias

Consumers

procrastinate

Claiming the EITC:

• Tax preparation companies

would like to reduce this

bias

Claiming rebates:

• Retailers would like

to exploit this bias

Barr, Mullainathan and Shafir (2013), “The Case for Behaviorally Informed Regulation”

24

The New Yorker, February 1, 2010 25





New policy tools



26

Behaviorally Informed Regulation:

The Case of Disclosure

Reduce Information Asymmetries

Reduce Search Costs

Traditional Rationale for Disclosure Requirements

27

S&P 500 Index Fund Experiment

Subjects allocate a

hypothetical $10K across

four real S&P 500 Index

Funds

All subject groups are more

financially literate than the

population of American

investors

Harvard undergraduates

Wharton MBAs

Harvard non-faculty staff

28

Experimental Conditions

Control

Subjects receive only four prospectuses

Prospectuses are often the only information investors receive from companies

Fees transparency treatment

Eliminate search costs by also distributing fee summary sheet (repeats information in prospectus)

Returns treatment

Highlight extraneous information by distributing summary of funds’ annualized returns since inception (repeats information in prospectus)

29

Portfolio Fees of Control Group:

Student Experiment

$421 $431

$309

$349

$389

$429

$469

$509

$549

$589

MBA College

Control

Minimum

Possible

Fee

Maximum

Possible

Fee

$443: average fee with random fund allocation

0% of College Controls

put all funds in

minimum-fee fund

6% of MBA Controls

put all funds in

minimum-fee fund


30

Fee Treatment Effect on Portfolio Fees:

Student Experiment

$421 $431

$366

$410

$309

$349

$389

$429

$469

$509

$549

$589

MBA College

Control

Fee treatment

19% of MBA Controls

put all funds in

minimum-fee fund

10% of College Controls

put all funds in

minimum-fee fund


31

Returns Treatment Effect on Fees:

Student Experiment

$421 $431

$440

$486

$309

$349

$389

$429

$469

$509

$549

$589

MBA College

Control

Returns treatment


32

Lessons on Disclosure

Lesson 1: Consumers are inattentive to relevant

information when it is not salient

Lesson 2: Consumers do respond to relevant

information when it is salient and easy to find

Although doesn’t necessarily result in optimal decision-

making

Lesson 3: Consumers also respond to irrelevant

information when it is salient

33

Making Disclosure More Effective

• Do the calculations C

• Translate information to personal objectives O

• Provide relative comparisons R

• Expand important outcomes E

Source: “Designing Better Energy Metrics for Consumers” (Larrick, Soll and Keeney, 2016)

34

C: Do the Calculations 35

O: Translate Information to Personal

Objectives 36

R: Provide Relative Comparison

Source: https://www.canada.ca/en/financial-consumer-agency/services/loans/payday-loans.html

37

E: Expand the Outcome 38

Evidence on the Impact of Behaviorally

Informed Disclosure 39

Behaviorally Informed Disclosure on

Future Payday Loan Utilization

Source: Bertrand and Morse (2011)

-5,1% -5,0%

-6%

-5%

-4%

-3%

-2%

-1%

0%

Relative comparison

Dollar calculation over time

Fu

ture

pay

day

loan

utiliza

tion

re

lative to c

ontr

ol gro

up

Disclosure Treatment

40

Increase in the Likelihood of Choosing the

Lowest Cost Option

(Relative to Control Group)

Loan

Choice

Savings Account

Choice

Comparison Table

5 options +25 pp +7-18 pp

10 options No effect +12 pp

Cost in pesos vs.

percentages +8 pp +4 pp

The Effect of Behaviorally Informed

Disclosure: A Study in Peru and Mexico

Source: Gine, Cueller and Mazer, “Information Disclosure and Demand Elasticity of Financial Products,” 2017

41





New policy tools



42


Products

Standardized “plain

vanilla” products

Reduces complexity of

product comparison

Facilitates disclosure of

relevant

characteristics/prices

Facilitate price competition

Prohibitions on risky

products/product features

Fee caps

43


Process

Choice architecture

The design of

environment and

context in which

choices are made

44


Process

“Smart” defaults

“Smart” choice menus

Fewer options

Tiered options

Facilitate comparison

Decision support tools

Decision timing

Decision verification

45


People

Consumers:

Require demonstration of

understanding before

allowing certain

transactions involving risky

products/large stakes

Financial services industry

employees:

Conduct standards (e.g.,

suitability requirements,

fiduciary standard)

46

A lot of our policy models

traditionally are based on a

rather naïve understanding of

what drives behavior. But if you

have a more intelligent, nuanced

account of how people make

decisions, you can design policy

that is more effective, less costly,

and makes life easier for most

citizens.

—David Halpern, Director of the

UK Behavioural Insights Team

Can We Improve Government? 47

brigitte madrian harvard university

Documents