How intra‐ and interfirm agglomeration affect new‐unit geographic distance decisions of multiunit firms

• Author: Hyun‐Soo Woo,Luiz Mesquita,Albert Cannella
• Date: 01 November 2019
• Published date: 01 November 2019
• DOI: http://doi.org/10.1002/smj.3070

RESEARCH ARTICLE

How intra- and interfirm agglomeration affect new-

unit geographic distance decisions of multiunit firms

Hyun-Soo Woo

1

| Albert Cannella Jr

2

| Luiz Mesquita

3

1

School of Business Administration,

University of Mississippi, Oxford,

Mississippi

2

Mays Business School, Texas A&M

University, College Station, Texas

3

W. P. Carey School of Business, Arizona

State University, Tempe, Arizona

Correspondence

Hyun-Soo Woo, School of Business

Administration, University of Mississippi,

87 Galtney Lott Plaza, University, Oxford,

MS 38677.

Email: hwoo@bus.olemiss.edu

Abstract

Research Summary:Prior agglomeration research takes a

competitive-level view, where any incumbent is consid-

ered as a rival by a new entrant in the same geographic

market. Our study suggests an alternative corporate-level

view, where entry by multiunit firms must consider sister

units as well as rivals in the market. Theorizing about a

sharing mechanism between sister units distinct from a

spillover mechanism between rivals, we expect that multi-

unit firms locate new units nearer sister vis-à-vis rival

units and that the size, quality, and organizational form of

a new unit moderate these associations. Finally, we predict

that multiunit firms establish new units distant from same-

brand and same-market-segment sister units. We find

robust empirical support from the geographic distance

decisions of 10 multiunit hotel firms in 20 U.S. cities.

Managerial Summary:Where should multiunit firms

(e.g., fast-food chains, hotels) locate new business units

relative to others? Current competitive-level-strategy per-

spective argues new units should locate “far-from-others”

if they have superior capabilities (to avoid being imitated)

and “near-others”if inferior ones (to better imitate others).

We instead examine this phenomenon from a corporate-

level-strategy perspective, regarding locally available syn-

ergies sister units can garner to better compete. We argue

that new units locate nearer sister (i.e., business-units of

same-parent firms) than rival units; that this effect is stron-

ger if new units are larger, have better quality, and are

company-operated rather than franchised; and that this

base effect is weaker when sisters belong to same- rather

Received: 26 May 2017 Revised: 23 May 2019 Accepted: 6 June 2019 Published on: 5 August 2019

DOI: 10.1002/smj.3070

Strat Mgmt J. 2019;40:1757–1790. wileyonlinelibrary.com/journal/smj © 2019 John Wiley & Sons, Ltd. 1757

than different-brand companies of the same parent-firm.

We provide supporting empirical evidence from the hotel

industry.

KEYWORDS

agglomeration externalities, geographic distance decision, interfirm

agglomeration, intrafirm agglomeration, multiunit firms

1|INTRODUCTION

Why do firms in the same industry locate near each other?

1

This phenomenon, known as agglomera-

tion, has received considerable attention from strategy scholars (e.g., McCann & Folta, 2008). Prior

agglomeration research has implicitly taken a competitive-level view, where any incumbent firm is

considered to be a rival by a new entrant in the same geographic market. This view, which we shall

refer to as interfirm agglomeration (Ramos & Shaver, 2013; Rawley & Seamans, 2015), explains

geographic distance decisions as driven by a spillover mechanism: if a new entrant perceives that it

will benefit from incumbents via access to specialized know-how or scale-efficient resources made

available in the local agglomerated area, it will locate nearby (Kalnins & Chung, 2004). In turn, if a

new entrant expects to create such benefits, it will instead choose to locate far from incumbents in

order to prevent valuable competitive resources from spilling over to local rivals (Shaver &

Flyer, 2000).

The competitive-level view provides a vital theoretical lens to explain the geographic colocation

of business units, but it is limited in two main ways. First, it presumes agglomeration among rivals,

although anecdotal observation suggests that agglomerated areas can also include business units of

same parent firms (henceforth, “sister”units). This latter concept, which we refer to as intrafirm

agglomeration,remains relatively unexplored. A large number of industries—particularly those

where goods are consumed locally, such as fast-food, health care, hotels, banking, drugstores, and

many others—are characterized by a mix of sister and rival units located near one another

(e.g., Garvin & Levesque, 2008; Kalnins, 2004a; Ramos & Shaver, 2013; Rawley & Seamans,

2015). Second, the spillover mechanism that the competitive-level view relies upon to explain geo-

graphic colocation choices by rivals does not explain geographic colocation between sister units.

Business units from same-parent firms are not simply independent rivals, and often cooperate in

unique ways to maximize benefits, both to themselves locally, and to their parent firms. If a

competitive-level view posits that spillovers of agglomeration benefits to others in the same area are

mostly undesirable, the view changes when the focus of interest is sister units. Unlike rivals, sister

units actively promote internal know-how sharing as well as local resource pooling to increase scale

efficiencies and specialization in key operational activities (Alcácer & Zhao, 2012; Golden & Ma,

2003). Hence, we currently lack a deep understanding of the internal sharing versus the external spill-

over mechanisms behind the respective colocation of sister and rival units.

1

Our reference to the interfirm (or “competitive-view”) agglomeration focuses on the geographic colocation of firms in a

single industry and the benefits created thereby (i.e., Marshallian agglomeration). Another literature considers geographic

colocation between firms in different industries and at different production stages (i.e., Jacobian agglomeration) and is based

on a theoretical heritage and set of premises that are beyond the scope of our study. We refer interested readers to McCann and

Folta (2008) and van der Panne and van Beers (2006) for comprehensive comparisons of these two agglomeration literatures.

1758 WOO ET AL.

To fill this gap, we build a corporate-level view of agglomeration. The corporate-level view pro-

vides a unique logic to explain the agglomeration of sister units, whereby the geographic distance

decisions by same-parent firms are driven by a mechanism of internal resource development and

sharing. At the local geographic level, this mechanism relates to internal allocations of know-how

(e.g., coordination of local market and customer information) and operational assets (e.g., transfer of

local tasks or resources), as well as efforts to coordinate competitive initiatives (e.g., protecting cus-

tomer overflow to rivals, and serving more customers by sister units). As the internal sharing occurs

directly between local sister units, the advantages are less likely to spill over to nearby rivals. This

corporate-level view along with its underlying internal sharing mechanism help us predict—as a gen-

eral effect—that when a multiunit firm opens a new establishment in the same geographic market

where both rival and sister units already operate, it will locate the new unit closer to sister than to

rival units.

The corporate-level view also predicts unique colocation effects that contrast with those we would

observe from a competitive-level view. In the competitive-level view, early research argues that net

contributor firms (hereafter contributor)—those that create more spillovers than they absorb—locate

geographically farther from incumbents, regardless of their being sisters or rivals, while net benefi-

ciary firms (hereafter beneficiary)—those that absorb more spillovers than they create—locate closer.

In contrast, from a corporate-level view, firms will locate contributor units closer to sister than to

rival units in order to more easily share internal resource benefits. Further, the corporate-level view

suggests that rivalry can exist among sister units, and we expect that in the same geographic area,

same-brand and same-market segment sister units suffer from greater rivalry relative to different-

brand or different-market segment sister units. This corporate-level effect, hence, helps us predict that

multiunit firms will locate new establishments farther from same-brand or same-market-segment sis-

ter units than from different-brand or different-market-segment sister units, to manage these negative

intrafirm agglomeration effects.

To empirically validate our theoretical model on intra- versus interfirm agglomeration effects, we

tracked 10 large multiunit firms in the U.S. hotel industry for 23 years (1991–2013), comparing the

geographic distances of 1,649 new units they opened vis-à-vis incumbent rivals and sister units

already operating in the same geographic area. We also consider that multiunit firms vary with regard

to using same- or different brands and aiming at same- or different-market segments. Our data ana-

lyses provide robust support to our hypotheses. We join a nascent effort by scholars to examine loca-

tion advantages based on corporate-level effects; although to our knowledge, ours is the first study to

assess intrafirm agglomeration effects to predict the colocation of new establishments by multiunit

firms vis-à-vis both incumbent rivals and sister units in the same agglomerated market. In the discus-

sion section, we expand on these theoretical contributions, and discuss further implications for future

research and management practice.

2|A CORPORATE-LEVEL VIEW OF GEOGRAPHIC

COLOCATION

2.1 |Competitive-level view versus corporate-level view

Specialized industries—where incumbents produce similar products—often involve rivals locating

nearby due to agglomeration externalities—that is, benefits a firm attains by locating near rivals

(Marshall, 2013; McCann & Folta, 2008; Wang, Madhok, & Li, 2014). The competitive-level view

of agglomeration considers other local firms as rivals and suggests that colocation decisions are

WOO ET AL.1759