Presentation Title

Do Slot Machines Cause Bankruptcy? Spatial Difference in Difference Evidence

Panel Title

Session 2-3-B: True or False: “The Real Problem Is Not Whether Machines Think—But Whether Men Do” (B.F. Skinner)

Location

The Mirage Hotel & Casino, Las Vegas, Nevada

Start Date

8-6-2016 2:00 PM

End Date

8-6-2016 1:30 PM

Abstract

This is the first study to provide causal evidence on the impact of slot machines on individual consumer bankruptcy filings, using recent econometric techniques designed to identify causality. We exploit very detailed administrative data from Alberta, Canada, on the location of the universe of slot machine locations, and administrative data on the location of universe of individual bankruptcy filers.

Specifically, we compare the impacts on neighborhood bankruptcies of a bar or restaurant with a large amount of gambling removing its slot machines, compared to a bar or restaurant with a small amount of gambling removing its slot machines. We exploit our ability to observe the total dollar magnitude of gambling at every bar and restaurant.

In order to control for unobservable neighborhood characteristics, we use the spatial difference in difference econometric methodology. This methodology differences out unobservable neighborhood characteristics by comparing bankruptcy filers that are fractions of a kilometer (radii of 0,25 km, 0.5 km and 0.75 km) from the specific bar or restaurant, compared to bankruptcy filers slightly further away (radii of 1.5 km, 2.0 km and 2.5 km).

We find that a 1% increase in total dollars gambled in the year prior to the removal of slots from the specific bar or restaurant, causes a 1% to 3% reduction in bankruptcy filers two years after the closure of that slot machine location. These effects are much larger for very close neighbors of the slots location (0.25km) compared to neighbors slightly further away (0.5 km and 0.75 km).

Streaming Media

Do Slot Machines Cause Bankruptcy? Spatial Difference in Difference Evidence

Keywords

Slot Machines, Bankruptcy, Causality

Disciplines

Behavioral Economics | Economics | Statistics and Probability | Substance Abuse and Addiction

Comments

Audio recording of this presentation is attached as a downloadable MP3 audio file, 33.1 MB

23B_ScholnickBByunM_pres_DoSlot.mp3 (33175 kB)
Audio recording of presentation

Share

COinS
 
Jun 8th, 2:00 PM Jun 8th, 1:30 PM

Do Slot Machines Cause Bankruptcy? Spatial Difference in Difference Evidence

The Mirage Hotel & Casino, Las Vegas, Nevada

This is the first study to provide causal evidence on the impact of slot machines on individual consumer bankruptcy filings, using recent econometric techniques designed to identify causality. We exploit very detailed administrative data from Alberta, Canada, on the location of the universe of slot machine locations, and administrative data on the location of universe of individual bankruptcy filers.

Specifically, we compare the impacts on neighborhood bankruptcies of a bar or restaurant with a large amount of gambling removing its slot machines, compared to a bar or restaurant with a small amount of gambling removing its slot machines. We exploit our ability to observe the total dollar magnitude of gambling at every bar and restaurant.

In order to control for unobservable neighborhood characteristics, we use the spatial difference in difference econometric methodology. This methodology differences out unobservable neighborhood characteristics by comparing bankruptcy filers that are fractions of a kilometer (radii of 0,25 km, 0.5 km and 0.75 km) from the specific bar or restaurant, compared to bankruptcy filers slightly further away (radii of 1.5 km, 2.0 km and 2.5 km).

We find that a 1% increase in total dollars gambled in the year prior to the removal of slots from the specific bar or restaurant, causes a 1% to 3% reduction in bankruptcy filers two years after the closure of that slot machine location. These effects are much larger for very close neighbors of the slots location (0.25km) compared to neighbors slightly further away (0.5 km and 0.75 km).