Retail Agglomeration and Competition Externalities: Evidence from Openings and Closings of Multiline Department Stores in the U.S.

Abstract

From the perspective of an existing retailer, the optimal size of a cluster of retail activity represents a trade-off between the marginal increases in consumer attraction from another store against the depletion of the customer base caused by an additional competitor. We estimate opening and closing probabilities of multi-line department stores (“anchors”) as a function of preexisting anchors by type of anchor store (low-priced, mid-priced, or high-priced) using a bias-corrected probit model with county and year fixed effects. We find strong negative competitive effects of an additional same type but no effect on openings of anchors of another type. Supplementary materials for this article are available online.

KEY WORDS:

• Bias-corrected probit
• Closings
• Multi-line department stores
• Openings
• Shopping centers.

SUPPLEMENTARY MATERIALS

The supplementary materials contain seven appendices. Appendix 1 provides detailed classification descriptions and the criteria for defining anchor stores. Appendix 2 (supplementary to Table 2 and 3) contains robustness tests that are supplementary to Table 2 and 3 with an additional control for population. Appendix 3 (supplementary to Table 2) shows results of county-level openings based on fixed effect logit regressions with bias correction. Appendix 4 (supplementary to Table 3) shows results of county-level closings based on fixed effect logit regressions with bias correction. Appendix 5 (supplementary to Table 4) shows results of openings in growth, stable and decline Markets based on fixed effect logit regressions with bias correction. Appendix 6 (supplementary to Table 5) shows results of openings inside shopping centers based on fixed effect logit regressions with bias correction. Appendix 7 (supplementary to Table 6) shows results of freestanding anchors based on fixed effect logit regressions with bias correction.

ACKNOWLEDGMENTS

The authors thank discussants and participants at Tinbergen Institute Seminar, Amsterdam, Netherlands, October 2013; 8th Urban Economic Association Annual Meeting, Atlanta GA, November 2013; The Maastricht-NUS-MIT (MNM) Real Estate Finance and Investment Symposium, Maastricht, The Netherlands, September 2014; Reading University Seminar, England, 2013; Free University Seminar, 2013; North American Meetings of the Regional Science Association International (RSAI), Washington DC, November 2014; American Real Estate and Urban Economics Association (AREUEA)-ASSA conference, Boston MA, January 2015; UNC Charlotte Economics Seminar, Charlotte, February 2015. The authors thank Stephen Billings, Jenny Schuetz, Jan Rouwendal, Jos van Ommeren, Anupam Nanda, Gianluca Marcato, Janet Kohlhase, Xiaoming Li, Peng Liu, Seow Eng Ong, and Joseph Ooi.