Working Paper: NBER ID: w16739
Authors: Li Gan; Manuel A. Hernandez
Abstract: Agglomeration is a location pattern frequently observed in service industries such as hotels. This paper empirically examines if agglomeration facilitates tacit collusion in the lodging industry using a quarterly dataset of hotels that operated in rural areas across Texas between 2003 and 2005. We jointly model a price and occupancy rate equation under a switching regression model to endogenously identify a collusive and non-collusive regime. The estimation results indicate that clustered hotels have a higher probability of being in the potential collusive regime than isolated properties in the same town. The identification of a collusive regime is also consistent with other factors considered to affect the sustainability of collusion like cluster size, seasonality and firm size, and the results are robust to alternative cluster definitions.
Keywords: Agglomeration; Tacit Collusion; Lodging Industry
JEL Codes: C3; L13; L4
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| clustered hotels (Z30) | higher probability of being in collusive regime (D79) |
| collusive regime (D70) | higher prices (D49) |
| collusive regime (D70) | lower occupancy rates (R21) |
| collusive regime (D70) | lower dispersion in prices (D39) |
| collusive regime (D70) | lower dispersion in occupancy rates (R21) |
| agglomeration (R11) | facilitates tacit collusion (D43) |