Negative Options and the Limits of Notice Jens Grossklags Ph.D. - - PowerPoint PPT Presentation

negative options and the limits of notice
SMART_READER_LITE
LIVE PREVIEW

Negative Options and the Limits of Notice Jens Grossklags Ph.D. - - PowerPoint PPT Presentation

Negative Options and the Limits of Notice Jens Grossklags Ph.D. Candidate, School of Information, UC Berkeley Deirdre Mulligan, Clinical Professor Chris Jay Hoofnagle, JD Maryanne McCormick, JD Nathaniel Good, Ph.D. Candidate Samuelson Law,


slide-1
SLIDE 1

FTC Neg. Options Workshop, January 25, 2007

Negative Options and the Limits of Notice

Jens Grossklags Ph.D. Candidate, School of Information, UC Berkeley Deirdre Mulligan, Clinical Professor Chris Jay Hoofnagle, JD Maryanne McCormick, JD Nathaniel Good, Ph.D. Candidate Samuelson Law, Technology & Public Policy Clinic, U.C. Berkeley Boalt Hall School of Law

slide-2
SLIDE 2

2

FTC Neg. Options Workshop, January 25, 2007

Overview

I.

Short notices produce better outcomes

∴Notices should be short, conspicuous

II.

But regret is still present, and framing can manipulate behavior detrimentally

∴Notice should include total cost

III.

Delayed consequences, inertia creates suboptimal decisions

∴Easy cancellation necessary

slide-3
SLIDE 3

3

FTC Neg. Options Workshop, January 25, 2007

  • Subjects were presented with an installation dialog

for three potentially harmful, popular programs

  • We removed brand information from programs
  • Three subexperiments:

1.

Standard EULA (64 subjects)

2.

Standard EULA + short notice before installation (80 subjects)

3.

Standard EULA + short notice after installation (78 subjects)

  • I. Experiment: EULA v. Short Notice
slide-4
SLIDE 4

4

FTC Neg. Options Workshop, January 25, 2007

Subexperiment I: Only Standard EULA

slide-5
SLIDE 5

5

FTC Neg. Options Workshop, January 25, 2007

Terms Are Not Read

  • Majority does not read

EULA information Median ≈ 45 sec Time required to pass through EULA is more than 10 min per program

Graph of reading time for individuals that installed programs (in seconds)

200 400 600 800 1000 1200 1400 1600 1800 10 20 30 40 50 60 Number of subjects (ordered by reading time) Reading time (in sec) Program X Program Y Program Z

slide-6
SLIDE 6

6

FTC Neg. Options Workshop, January 25, 2007

EULA Only: Most Install Potentially Harmful Programs

Most people install all three programs

– Programs X: 70.3%; Y: 90.6; Z: 85.9%

Installations are regretted

– Only few consumers would keep potentially harmful

programs installed after consultation; Programs X: 2.2%; Y: 62.1%; Z: 18.2%

Consumers do care

– Some differentiate between programs based on standard

EULA

– Strong response to EULA summary – In context of privacy and spyware: Motivates use of short

notices (favored model by FTC, EU, companies etc.)

slide-7
SLIDE 7

7

FTC Neg. Options Workshop, January 25, 2007

Subexperiments II & III: Short Notices Added

slide-8
SLIDE 8

8

FTC Neg. Options Workshop, January 25, 2007

Short Notices Reduce Bad Decisions

Users install fewer

potentially harmful programs with Pre- and Post- Notice

Many users keep less

risky program

Post Pre

90.6% 71.8% 29.5% 22.5% 70.3% 67.5% 85.9% 30.8% 23.8% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% EULA only EULA and Pre notice EULA and Post notice Program X Program Y Program Z

Programs installed (in percent)

slide-9
SLIDE 9

9

FTC Neg. Options Workshop, January 25, 2007

Short Notices Are Better, But Consumers Still Regret Outcomes

Regret significantly lower in short notice

subexperiments, but still high overall

Up to 70% still regret decision to install in Pre-

notice subexperiment, 78% in Post

Experiment shows that notices can be largely

effective

slide-10
SLIDE 10

10

FTC Neg. Options Workshop, January 25, 2007

Short Notices: Regret & Other Problems Endure…

The terms of the bargain are not fully understood Optimism colors estimates of risk, ability to make

decisions, and interpretation of ambiguous terms

Excitement often causes them not to read terms Even the best notice practices have some limits…

slide-11
SLIDE 11

11

FTC Neg. Options Workshop, January 25, 2007

  • II. Limits of Disclosure

“Hidden costs” cause suboptimal decision making Even costs framed as “fees” affect decisions – Experiment in eBay auction setting showed that

increasing shipping costs while lowering opening price attracts higher number of bidders and increases sellers’ revenue (Hossain & Morgan)

∴Disclosure should include total cost

slide-12
SLIDE 12

12

FTC Neg. Options Workshop, January 25, 2007

  • III. Delayed Consequences

Complicate Decisions

Because negative options involve charges over time,

individuals may depart from perfect rational choice

Models developed in psychology and economics

describe and explain human behavior under these circumstances

slide-13
SLIDE 13

13

FTC Neg. Options Workshop, January 25, 2007

Effects of Delayed Consequences

Overwhelming incentives to sign up

– Individuals are enticed by immediate benefits

Driven by opt-out enrollment; free-to-pay conversions

– Reinforced by limited information/awareness about total costs

Consumer intent easily can be frustrated

– Procrastination of cancellation decision; inertia (or status quo

bias)

– Reinforced by burdensome cancellation procedures

slide-14
SLIDE 14

14

FTC Neg. Options Workshop, January 25, 2007

Recommendations

Short, conspicuous notices will help

– Consider that pool of consumers is heterogeneous

“Total cost of ownership” information desirable

– Disclosures should prevent sellers from hiding costs or

moving fees to the “back of the product”

“Mutuality” in cancellation procedures

– Protect consumer intent with cancellation procedures that are

as facile as enrollment procedures