1 of 41

The Tragedy of your Upstairs Neighbors:

When is the Home-Sharing Externality Internalized?

Apostolos Filippas John Horton

NYU Stern NYU Stern

1

2 of 41

Rapid rise in "home-sharing"

  • Platforms enable owners to rent out properties to short-term guests
    • e.g., Airbnb, HomeAway, VRBO, CouchSurfing
                  • ​
  • Rapid growth
    • 100+ million nights will be booked on Airbnb in 2017
    • valued at $31 billion
    • Zervas et al. 2014, Cusumano 2015, Farronato & Fradkin 2017
                  • ​
  • Mostly viewed as beneficial development
    • increased variety, underutilized assets put to use
    • reduced costs for guests, income for asset owners
    • Sundararajan 2016, Einav et al. 2016, Horton and Zeckhauser 2016

2

3 of 41

Policy issues raised by home-sharing

  • Home-sharing platforms facilitate racial discrimination
    • Edelman et al. 2017
                  • ​
                  • ​
  • Home-sharing raises long-term rents by removing supply
    • Sheppard & Udell 2016
                  • ​
                  • ​
  • Home-sharing directly imposes costs on neighbors
    • This paper

3

`

4 of 41

4

Neighbors

​

Guest

​

Owner

​

$

$

$

$

5 of 41

5

6 of 41

Home-sharing as a negative externality

  • Hosts impose a cost on their neighbors
    • hosts get the money, guests get the noise/risk/parties/...
                  • ​
                  • ​
  • Similar to classic cases of un-internalized externalities
    • pollution, overfishing, speeding while driving
                  • ​
                  • ​
  • There is "too much" of the activity in equilibrium
    • in practice, negative externalities solved with public policy

6

7 of 41

What are some public policy options?

  • Common policy approaches likely to be ineffectual
    • Coasian bargaining
    • Charge per unit of externality, marketable hosting permits

​

  • Another approach:
    • Allocate the decision right to hosts and see what happens

7

8 of 41

Research Questions

  1. Who should make home-sharing hosting decisions?

​

8

9 of 41

What happens when the decision is made by:

9

Individual

tenants

Building

owners

City

government

Benevolent

social planner

10 of 41

Our analytic approach

  1. Allocate hosting decision rights to: {individuals, building owners, city, social planner}
                  • ​
                  • ​
  2. Determine the resulting market equilibrium
                  • ​
                  • ​
                  • ​
  3. Characterize the welfare properties of that equilibrium�

10

11 of 41

Some definitions

  • The going price for a home-sharing stay is
    • price is endogenous: decreasing in number of hosts
                  • ​
  • Tenant/host has hosting cost
                  • ​
  • The tenant/host has neighbors in the building
                  • ​
  • A guest imposes a cost of on every neighbor
                  • ​
  • An additional listing is (socially) efficient so long as

11

12 of 41

The four regimes

Set building-wide policies Care only about profits from long-term rents

Host if

Centrally adjusts supply Cares only about tenants

Centrally adjusts supply Cares about tenants and guests

​

12

13 of 41

Market equilibrating process

Move to the "right" building

Host if

Set policies

Price updated

13

14 of 41

Market equilibrating process

Move to the "right" building

Host if

Set policies

Price updated

Costlessly (we will relax this later)

14

15 of 41

Tenants-Decide (TD) regime

​

  • Laissez-faire
    • tenants are free to decide whether to host or not
                  • ​
                  • ​
  • The "marginal" tenant who hosts, has hosting cost equal to the market price
                  • ​
                  • ​
  • This is inefficient → there is too much hosting.

15

16 of 41

Tenants-Decide regime

​

16

  • Inefficiently high amount of hosting
        • ​
  • Tenant welfare ↘ (potentially)

17 of 41

TD regime - externalities

17

  • Externalities are not internalized

18 of 41

"City decides" (CD) regime

  • The City decides on a home-sharing quantity
                  • ​
                  • ​
  • The City only cares about residents’ welfare
    • behaves like a monopolist
                  • ​
                  • ​
  • There is too little hosting

18

19 of 41

"City decides" regime

​

19

​

  • Host welfare ↗ (optimal)
          • ​
  • Social welfare ↗ (but not optimal)

20 of 41

"Building owners" (BD) decide

  • Building owners
    • only care about profits from long-term rentals
    • set a blanket policy for the entire building
                  • ​
  • In equilibrium:
    • owners are indifferent between costless policies → no premium "no policy arbitrage"
                  • ​
    • marginal host indifferent between hosting bldg and non-hosting bldg on basis of hosting costs and benefits (rents are the same), and so:

​

20

21 of 41

"Building owners" decide

​

21

  • Socially optimal amount of hosting
        • ​
  • Tenant welfare ↗ (always)

22 of 41

BD regime - externalities

22

  • Externalities internalized

23 of 41

Research Questions

  • Who should make home-sharing hosting decisions?
    • Building Owners
                  • ​
  • Does our "no policy arbitrage" prediction hold?

​

23

24 of 41

"No policy arbitrage"

  • In equilibrium, building owners should be indifferent over costless policies
                  • ​
  • Problem: home-sharing is relatively nascent
    • data from existing rental markets unlikely to offer a compelling test
                  • ​
  • Solution: test this prediction using subletting decisions
    • Conceptually similar
    • Slight administrative cost
    • Large externality costs to neighbors & prospective renters

24

25 of 41

Data

25

  • ~21,000 apartment listings in NYC
  • Attributes of each listing (# beds, amenities, ... )
  • Whether each listing allows for subletting

26 of 41

Assessing the "No policy arbitrage" prediction

26

27 of 41

Assessing the "No policy arbitrage" prediction

27

Counterfactual

28 of 41

Research Questions

  • Who should make home-sharing hosting decisions?
    • Building Owners
                  • ​
  • Does our model’s "no policy arbitrage" prediction hold?
    • Strong evidence that it does
                  • ​
  • What are the convergence properties of a market operating under the BD policy regime?

28

29 of 41

Converging to the BD equilibrium

  • Our equilibrium analysis assumes tenant sorting
    • tenants have to move into the “right” building
                  • ​
                  • ​
  • Problem: convergence not guaranteed
  • Problem: excessive amount of tenant sorting required
  • Problem: moving costs & tenant "lock-in"
                  • ​
                  • ​
  • Solution: use an agent-based model

29

30 of 41

Convergence

30

31 of 41

Moving Costs

31

32 of 41

Within-building type correlation

32

33 of 41

Research Questions

  • Who should make home-sharing hosting decisions?
    • Building Owners
                  • ​
  • Does our model’s "no policy arbitrage" prediction hold?
    • Strong evidence that it does
                  • ​
  • What are the convergence properties of a market operating under the BD policy regime?
    • Good convergence properties
    • Moving costs matter (negatively)
    • Within-building correlation matters (positively)

33

34 of 41

Concluding thoughts

  • Delegating home-sharing decision rights to building owners seems to be a good idea
    • socially optimal, externalities internalized
    • market-based policy: information-light, self-adjusting (Tirole, 2015)
                  • ​
  • If you have to be in the CD regime, assign home-sharing licenses to buildings (not to tenants)

34

35 of 41

Concluding thoughts

New problems that home-sharing platforms create need to be addressed before their benefits are to be fully realized

  1. Within the platform
    1. Discrimination
    2. Other design issues
          • ​
          • ​
  2. Outside the platform
    • Houses off the rental market
    • Negative externalities

35

36 of 41

Concluding thoughts

New problems that home-sharing platforms create need to be addressed before their benefits are to be fully realized

  • Within the platform
    • Discrimination
    • Other design issues
          • ​
          • ​
  • Outside the platform
    • Houses off the rental market
    • Negative externalities

Doable!

36

37 of 41

Thank you

Paper Title: The Tragedy of Your Upstairs Neighbors: Is the Home-Sharing Externality Internalized?

Authors: Apostolos Filippas and John Horton

Draft: http://www.john-joseph-horton.com/papers/airbnb.pdf

37

38 of 41

Backup Slides

38

39 of 41

Benefits of Home-Sharing

39

Product choice set is expanded

Underutilized resources put to use

40 of 41

Tenants’ building choice

40

41 of 41

41