An Aggregate Generalized Nested Logit Model of Consumer Choices: An Application to the Lodging Industry

45 Pages Posted: 7 Oct 2007

See all articles by Sriram Venkataraman

Sriram Venkataraman

University of North Carolina - Chapel Hill, Kenan-Flagler Business School

Vrinda Kadiyali

Cornell University - Samuel Curtis Johnson Graduate School of Management

Date Written: October 2005

Abstract

In this paper, using aggregate data, we demonstrate the ability of the generalized nested logit (Wen and Koppelman, 2000; GNL henceforth) to better capture consumer choice under conditions where consumer tradeoffs among choice items is not ex-ante obvious to the researcher, and data on attributes of consumer choice are incomplete. We extend existing GNL models by using more readily available aggregate data (rather than individual data) while accounting for consumer heterogeneity and endogeneity of firm-choice variables. The empirical application is to the lodging (hotel) industry. The industry classifies properties on the basis of price tiers; it also recognizes that consumers appear to have two idea points (of downtown and airport) for location. However, it appears possible that a consumer might see a property in the same location but a different price tier as a closer substitute than a property of the same price tier at a different location. Hence for this industry, an ex-ante nesting structure based on price tiers or location alone might not capture the complexities of consumer choices. We find that GNL provides a better fit to these data than aggregate logit or aggregate probit. Our results provide managerially useful insights into who might comprise competition for any hotel, i.e. is it a nearby property of the same/different quality tier or is it a distant property of the same/different quality tier. We also briefly discuss the implications of consumer choices for firm profitability by estimating a supply-side model.

Keywords: consumer choice models, aggregate data, generalized nested logit

Suggested Citation

Venkataraman, Sriram and Kadiyali, Vrinda, An Aggregate Generalized Nested Logit Model of Consumer Choices: An Application to the Lodging Industry (October 2005). Johnson School Research Paper Series No. 12-07, Available at SSRN: https://ssrn.com/abstract=1019534 or http://dx.doi.org/10.2139/ssrn.1019534

Sriram Venkataraman

University of North Carolina - Chapel Hill, Kenan-Flagler Business School ( email )

McColl Building
Chapel Hill, NC 27599-3490
United States
9199620992 (Phone)

HOME PAGE: http://https://www.kenan-flagler.unc.edu/faculty/directory/sriram-venkataraman/

Vrinda Kadiyali (Contact Author)

Cornell University - Samuel Curtis Johnson Graduate School of Management ( email )

Ithaca, NY 14853
United States
607-255-1985 (Phone)
607-254-4590 (Fax)

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
403
Abstract Views
2,076
Rank
134,085
PlumX Metrics