Search ▸ Communication to the City Council
CRT 2016 #50 · Communication to the City Council · Sep 12 2016
a report from Councillor Craig Kelley, Chair of the Public Safety Committee, Mayor E. Denise Simmons, Co-Chair of the Housing Committee, Vice Mayor Marc C. McGovern, Co-Chair of the Housing Committee, for a joint public hearing held on August 3, 2016 to discuss the operations and prevalence of short-term rental units in Cambridge, and further develop the information and opinions gathered during the initial hearing on July 19, 2016
Racial Discrimination in the Sharing Economy:
Evidence from a Field Experiment
Ben Edelman, Michael Luca, and Dan Svirsky
Booking at Marriott
Booking at Airbnb
Searching on Airbnb
Anecdotal Evidence
But the plural
of anecdote
isn’t data…
Testing for Discrimination on Airbnb
• Field experiment:
• Send out roughly 6,400 requests for availability.
• Vary name of guest. Hold all else constant.
• Cities:
• LA
• Washington, DC
• St Louis
• Dallas
• Baltimore
• Draw 6,400 properties with availability one month out.
• Randomly assign each host to get a message from either an
African-American guest or a White guest
• Send message inquiring about availability. Vary only guest’s
name.
Guest Names
White Names
Allison Sullivan
Brad Walsh
Anne Murphy
Brent Baker
Kristen Sullivan
Brett Walsh
Laurie Ryan
Greg O’Brien
Meredith O’Brien
Todd McCarthy
African-American Names
Lakisha Jones
Darnell Jackson
Latonya Robinson
Jamal Jones
Latoya Williams
Jermaine Jones
Tamika Williams
Rasheed Jackson
Tanisha Jackson
Tyrone Robinson
A host’s
perspective
Main Results
•
Guests with distinctively white names got a
Yes response 50% of the time.
•
Guests with distinctively African-American
got a Yes response 42% of the time.
Difference is persistent across:
•
Cities
•
Race of host
•
Price of property
•
Whether host lives at the property
Are Black hosts different?
Yes Response
Guest is Black
-0.08***
(0.02)
Host is Black
0.06
(0.04)
Guest, Host is Black
0.01
(0.05)
Constant
0.48***
(0.01)
Observations
6,235
Adjusted R2
0.007
Are shared properties different?
Yes Response
Guest is Black
-0.07***
(0.02)
Shared Property
0.00
(0.02)
Shared Property, Guest is Black
-0.02
(0.03)
Constant
0.49***
(0.01)
Observations
6,235
Adjusted R2
0.006
Large-scale hosts
Yes Response
Guest is Black
-0.08***
(0.02)
Host has Multiple Listings
0.10***
(0.02)
Host has Multiple Listings, Guest is Black
-0.00
(0.02)
Constant
0.46***
(0.01)
Observations
6,235
Adjusted R2
0.014
Does host age matter?
Yes Response
Guest is Black
-0.11***
(0.02)
Host Looks Young
-0.03
(0.02)
Host Looks Young,
Guest is Black
-0.01
(0.03)
Constant
0.50***
(0.01)
Observations
6,235
Adjusted R2
0.011
Role of neighborhood
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0%
10% 20% 30% 40% 50% 60% 70% 80% 90%
White
Black
Names
Share of
black
population
in census
tract
Proportion
Yes
Responses
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
$10
$55
$68
$80
$95 $109 $129 $150 $199 $285
White
Black
Names
Proportion
Yes
Responses
Role of property price
Property
price per
night
Open Questions
•Why is this happening?
•What can make it better?
•Do positive reviews help?
•Will Airbnb’s efforts help?
•Who should be held responsible?
Other Open Regulatory Issues
• Taxes – Hotels pay (high) hotel taxes. Should Airbnb (valued
at $30 billion)? And if so – what about back taxes?
• Zoning – Hotels face strict limits in where they can place
their business.
• Safety
• Fire extinguishers, fire escapes, fire-proof bedding, fire-proof
furniture, backup lighting, deadbolts…
• Arbitration Clause
• Why shouldn’t a Massachusetts judge applying Massachusetts law
settle these disputes?
• Disclosures to regulators
• See San Francisco, New York