Top Banner
Leveraging Institutions for Competitive Advantage: How Firms Use the Business Media in Making Markets Mark Thomas Kennedy Marshall School Business University of Southern California Bridge Hall 303 3670 Trousdale Parkway Los Angeles, CA 90089-0808 Tel: (213) 821-5668 Fax: (213) 740-3582 [email protected] Edward J. Zajac Kellogg School of Management Northwestern University 2001 Sheridan Road Evanston, IL 60208 Tel: (847) 491*8272 [email protected] Submitted to 2006 Atlanta Competitive Advantage Conference February 3, 2006
57

Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

May 28, 2020

Download

Documents

dariahiddleston
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Leveraging Institutions for Competitive Advantage:

How Firms Use the Business Media in Making Markets

Mark Thomas KennedyMarshall School Business

University of Southern CaliforniaBridge Hall 303

3670 Trousdale ParkwayLos Angeles, CA 90089-0808

Tel: (213) 821-5668Fax: (213) 740-3582

[email protected]

Edward J. ZajacKellogg School of Management

Northwestern University2001 Sheridan RoadEvanston, IL 60208Tel: (847) 491*8272

[email protected]

Submitted to 2006 Atlanta Competitive Advantage Conference

February 3, 2006

DRAFT – Please do not copy or cite without the authors’ permission.

Page 2: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

2

Page 3: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

ABSTRACT

The sources of competitive advantage are many, and strategy and organizational

researchers have long sought to identify and understand what contributes to interfirm

differences in competitive outcomes. This study seeks to contribute to this literature by

focusing attention on the relationship between organizations and institutions. This

relationship has always defied simple description, since institutions limit organizations,

yet organizations leverage institutions. We propose and test a theory regarding how

institutions can be both limits to and levers for strategic organizational action.

Specifically, we propose that firms can generally leverage longstanding and stable

institutions to their advantage, but that more extreme leverage attempts will elicit

disadvantageous competitive outcomes as institutions respond in ways that preserve their

capacity to shape organizational environments. We draw on a field study of firms’

attempts to leverage their relationships with the business media to contextualize our

arguments and develop and test hypotheses suggesting that firms’ leverage attempts can

generate positive competitive outcomes, but generate negative outcomes when such

attempts violate norms of reporter independence. Quantitative analyses of media

coverage and firm performance outcomes support these hypotheses. We conclude by

discussing the implications of our approach and findings for future research on the role of

institutional environments and competitive advantage.

KEYWORDS

institutions, change, strategy, innovation, media, networks

Page 4: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

4

Page 5: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

INTRODUCTION

The relationship between organizations and institutions is a seeming contradiction: on the

one hand, institutions limit organizations while on the other, organizations leverage institutions.

Research has suggested that competing firms can increase their organizational survival chances

when linked to institutions (cultural, political, or economic) that convey legitimacy (Baum 1999;

Baum and Oliver 1992; Baum and Oliver 1991); similarly, it has been argued that firms can face

sanctions for following nonconformist strategies and/or organizational forms (Zuckerman 1999).

The existence and influence of institutions ranging from property rights to industry associations

to operating procedures is often taken for granted by organizations (see especially Scott 1995),

generating conformity with little apparent resistance.

However, research also shows that organizations actively leverage institutions, gaining

the benefits of conformity by conforming to institutional demands for show only. This occurs

when organizations decouple their formally declared strategies and structures from substantive

activity (Meyer and Rowan 1977)—sometimes even by announcing but never actually

implementing organizational actions thought to be legitimate (Westphal and Zajac 1994; Zajac

and Westphal 1998; 1995). Of course, in some situations organizations may be able to go so far

as to disregard institutional demands without experiencing competitive disadvantages (Kraatz

and Zajac, 1996). This observed variation in the power of institutions on competing

organizations leaves us with an important and under-researched compound question: When are

institutions particularly relevant for competing organizations, and what mechanisms determine

whether relevant institutions are limiting versus leveraging aspects of competitive advantage?

We engage this question by building on work organizational sociologists have done to

explore institutions in society, particularly the role contradictions in institutions and society

Page 6: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

(Clemens and Cook 1999; Clemens 1997; Friedland and Alford 1991). Friedland and Alford

(1991: 103) observe that societies are built from multiple and potentially contradictory

institutions that each provide “a central logic … to organizations and individuals to elaborate”

what is legitimate. They use this to explain change by arguing that society’s institutional

relations can be transformed when individuals and organizations exploit their contradictions.

Clemens (1997) provides qualitative evidence of this process by showing that when exogenous

environmental shocks caused collisions between established institutions, savvy political actors

were actively elaborating the logic of key institutions to raise and resolve contradictions made

salient by demographic and technological changes. Picking up the pieces of institutions

effectively being broken apart by these changes, these ‘institutional entrepreneurs’ resourcefully

patched them together to create successor institutions that both built on and departed from their

predecessors, gradually replacing them.

This line of research inquiry has two valuable implications for our study. First, it shows

how institutions may be contradictory in their demands, and secondly, it provides evidence of the

interplay between the unusual event of dramatic institutional change and the active efforts of

organizations and individuals (see also Clemens and Cook 1999). However, we believe that

organizations typically face institutions that are more stable than those found in Clemens’ unique

research setting, and we believe further that the ultimate purpose of firms’ attempts to relate to

institutions is the desire to influence positively the competitive position of a firm. Thus, our

focus is on the following important and under-researched question: How can organizations, in

the pursuit of competitive advantage, take actions that effectively leverage stable and

longstanding institutions?

We address this question directly by theorizing and analyzing empirically how firms

leverage contradictions in the institutions of business journalism in the pursuit of self-serving 6

Page 7: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

conceptions of an emerging product market. The contradiction that we explore is the duality of

organizational attempts at leveraging institutions. More specifically, we use qualitative and

quantitative data to show evidence consistent with our notion that firms leverage the institutional

norms of the journalism profession to get their stories out to the public through editorially

controlled outlets that can communicate, validate and legitimate their strategies. Equally

important, however, we also find evidence consistent with our argument that firms that overreach

in their attempts to leverage institutions experience a negative backlash. As we will discuss, our

emphasis on the duality of leveraging institutions enables us to consider simultaneously both the

likely positive and negative competitive implications of such efforts.

FIELD STUDY AND HYPOTHESES

To understand the extent and effectiveness of the work firms do to shape public

perception about what they do to their advantage, we look at their attempts to place stories that

sell the claims they would like to make about an emerging market. There are three key features

of journalism that make it a great site for understanding the limits organizations’ ability to

leverage institutions for strategic change. First, the so-called free press is widely recognized as

an institution crucial to the transparency and information flows on which capitalist market

economies depend. Second, its independence norms create tensions for reporters who must

therefore convince sources to volunteer time and information without being able to control how

it will be used. Third, the competing risks of receiving favorable versus unfavorable coverage

create incentives for organizations to both give and withhold the information reporters need to do

their jobs. Thus, though each side tends to regard the other somewhat antagonistically, both

7

Page 8: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

sides need each other. This suggests a potential for reciprocal dealing that contradicts

journalism’s strong norms of reporter independence.

Moreover, especially in the context of new market formation, this potential for reciprocal

dealing suggests a route to establishing positions of competitive advantage that is both

understudied and clearly suggested by prior theory. For business organizations, establishing a

new product new market formation offers unique opportunities to develop long-lasting advantage

(Carpenter and Nakamoto 1989; Lieberman and Montgomery 1998). White’s work developing

the role of social comparison in market formation (White 1981; 2002) and Porac’s work showing

the importance of categories in structuring and shaping these comparisons (Porac and Rosa 1996;

Porac and Thomas 1990; Porac et al. 1995) combine to suggest we should take news coverage

seriously as a source of data about market formation, for it constitutes a large record of the

collective sensemaking about product similarities and differences. A notable start in that

direction was the Rosa et al. study of the market for minivans (1999) through a content analysis

of usage of various labels used for this new type of vehicle that was part truck, part car, part

station wagon, part van. That study showed media discourse matters to market formation: it

found that demand growth followed growth in use of the term ‘minivan’ in the press and the

stabilization of its meaning through all that accumulated usage. It did not analyze comparisons

among rivals, however, nor did it explore the structure of the category as it formed.

These themes were explored through direct participant observation of public relations

experts working with clients and reporters working with sources to prepare for the key annual

conference of a large media company specializing in business news. Based on early interviews,

field work focused on exploring the information flows between these parties and the incentives

driving them. Figure 1 shows the integrating framework distilled from following the information

flows among the various parties; this ‘diamond’ of four-way information flows shows each party 8

Page 9: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

giving and receiving something in return. Reducing field and interview notes about these flows

revealed a surprising reciprocity as a basic theme that was then developed further in follow-up

interviews and archival research designed to compare the conflicting and complementary views

of agents, reporters, sources and audiences.

[Figure 1 about here]

Source-Reporter Exchange

The information flows fueling business news coverage constitute an economy of

reciprocity (Polanyi [1944] 1985) driven by incentives that reward both cooperation and

defection between reporters and their sources at firms. Focusing on the reporter-source

exchanges, each side offers something the other side wants, and side can give or withhold what

the other side wants. On one side, firms seek coverage both to spread and legitimate stories about

the advantages of their products and services; reporters can give or withhold that coverage. On

the other side, reporters seek access to sources at firms both to add credibility and interest to their

stories about them and to meet stiff deadline pressures; sources can give or withhold this access.

This sets up incentives to try to trade access for coverage and vice versa. Firms stand to gain

from giving reporters the kind of open access that might cultivate a sense of social obligation and

reporters stand to gain from having privileged access to valued sources. As one reporter

described this dynamic, “Reporters have some sources that they care about over time. … You

take care of them.” Thus, reporters and sources have incentives to find a stable cooperative

equilibrium that sustains information flows between them.

9

Page 10: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

On the other hand, the norms of the journalism profession prevent reporters from making

any explicit bargains with sources for information from them, and this reporter independence is

what makes news coverage valuable to firms. As one marketer put it, “You need influencers that

will tell your story for you … [you] get other people to say that what you’ve got is the next big

thing worth being its own category.” The influence of news stories can only be maintained if

reporters sometimes use sources’ information against them: this means doing negative stories on

sources who would not have provided access had they known what was coming. As one reporter

put it: “I’m stupid if I don’t pay attention to stories from, say, IBM. Still, not every story IBM

wants to sell us is really interesting.” When the story is not interesting, the reporter risks losing

relevance if they do not clearly say so. From the source’s point of view, this feels like defection.

This type of defection may damage or even end the source-reporter relationship, however long-

or short-lived it may have been up to that point. Though reporters do weigh the costs of doing

negative stories, betraying a source may well pay when a negative story has the sort of

sensationalism that makes reporter reputations and sells news. Thus, reporters have incentives to

“get the dirt” and report it, and sources have incentives to carefully hide everything that could be

used against them, even if it means the sort of partial disclosure that would leave reporters

feeling lied to if they had know what was going on behind the scenes. Thus, both reporters and

sources have incentives to violate a cooperative equilibrium that would sustain open information

flows between them.

So reporters need access to sources who need coverage from them, and each side can give

or withhold what the other needs depending on how their exchanges go. This reciprocal

interdependence gives each side opportunities to influence the other by strategically giving or

withholding what the other side wants. For firms, supplying information to reporters eases

certain burdens of reporting, possibly leading to the coverage they desire. In the context of

10

Page 11: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

market-making, this comes to what one marketer described as “using” the press to sell their

market conception (Fligstein 1996). One marketer described the dynamic this way:

“We are using the press as our biggest medium. It’s not just reviews. We’ve been

working with outlets who can tell the story behind the product.”

While reporters may tolerate being used to some extent, protecting their independence creates

incentives for them to “push back” against more frequent attempts to use them. When “use”

becomes “abuse”, firms are therefore likely to see a negative impact of supplying information to

reporters. Without this push-back based in the institutional norms for reporter independence, the

journalism profession would lose its value. This, the structure of the institution holds features

that make it possible for firms to use it for competitive advantage while also protecting it from

overuse that could threaten it. This suggests a curvilinear relationship between coverage and

firms’ attempts to solicit it by supplying information to reporters, leading to our first hypothesis:

H1: (a) Supplying information (measured through press releases) to reporters and

analysts will be positively related to coverage, (b) but supplying too much

information to reporters will be negatively related to coverage.

But coverage is not so much an end to itself as it is a means to several important objectives:

a favorable position in the market, increased sales, and ultimately, survival. In terms of position,

one simple way firms think about competing is the relative share of attention they receive from

customers in their markets. Interviewees often referred to the concept of “mindshare” to capture

this, a word the Oxford English Dictionary defines as “consumer awareness of a particular

product or brand, esp. compared to the profile enjoyed by competitors' products.” That is, firms

11

Page 12: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

pursue coverage to realize the information advantages that come with being the firm most likely

to come to mind when people consider the market. As coverage benefits position (through

mindshare), it also holds the promise of increasing sales and the odds of long-term survival. One

problem with publicity-based marketing efforts, however, is that the evidence linking efforts to

secure coverage to coverage and its supposed benefits has been flimsy at best. This makes it the

following additional hypotheses valuable further tests:

H2: (a) Supplying information to reporters will be positively related to favorable

positions in the market (measured as mindshare in the emerging market –

explained below), (b) but supplying too much information to reporters will be

negatively related to favorable positions in the market.

H3: (a) Supplying information to reporters will be negatively related to the hazard rate of

market exit, (b) but supplying too much information will be positively related to

market exit.

H4: (a) Increasing attempts to leverage the press will initially be positively related to sales

but (b) overuse will have a negative impact on sales.

PR Agents and Reporters

The complexity of the source-reporter relationship leads many executives to hire public

relations (PR) agents, specialists in working with the press. PR agents collect approximately

$6.3B globally each year1 for assisting clients by bringing inside knowledge of newsrooms and

news personnel to the task of getting out a particular story. In the case of publicity-based

marketing efforts (as opposed to crisis management efforts), that story typically focuses on a new

1 XXX add source.12

Page 13: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

product, a new usage occasion, a new geography, or a new market created by some combination

of these. In effect, PR agents are brokers who help firms conform to reporters’ professional

norms for interacting with sources. Ironically, this submissive conformity to journalistic norms

disguises concerted attempts to undermine reporter independence by placing stories that are

tantamount to pre-packaged news stories. PR agents help their clients accomplish a decoupling

(Meyer and Rowan 1977) that is strategically motivated: they help them to conform journalistic

norms for reporter-source interactions while also helping them violate the norms for

independence. Since market making requires convincing reporters and the public to take

seriously departures from existing categories they are generally inclined to dismiss (Zuckerman

1999), the job may require skills some firms or executives do not have. If PR agents are bringing

skills that neither executives nor firm employees have then this clever non-conformist

conformity should create additional value, leading to the following hypotheses on the same four

dependent variables considered with coverage:

H5: (a) Initial increases in attempts to influence reporters in which PR agents are used

will be positively related to coverage, (b) but continued increases will be

positively related to coverage.

H6: (a) Initial increases in attempts to influence reporters in which PR agents are used

will be positively related to mindshare, (b) but continued increases will be

negatively related to mindshare.

H7: (a) Initial increases in attempts to influence reporters in which PR agents are used

will be negatively related to the hazard rate of market exit, (b) but continued

increases will be positively related to market exit.

13

Page 14: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

H8: (a) Initial increases in attempts to influence reporters in which PR agents are used

will be positively related to next period sales but (b) continued increases will be

negatively related to sales.

The Audience Connection

Given the vast volume of daily news and information, merely getting covered may not be

enough to help firms accomplish the aims of market-making. As mentioned, the job of market-

making requires persuading people to take a misfit product seriously as the founding instance of

what they should accept as a new product market category. Following density dependence

observed by organizational ecologists, the appearance of competition should initially increase the

life chances of entrant firms by lending a strength-in-numbers legitimacy to the new market

(Carroll and Hannan 1989), though, of course, continued entry should eventually create

competitive pressures that lead to increased exit. As firms are soliciting coverage, one way they

can attempt to “use” the news to influence audiences to take their claims seriously is to recognize

one or two rivals in their public information releases. Compared to only mentioning themselves,

then, this approach should have benefits to firms across all four dependent variables, leading to

the following additional hypotheses:

H9: Firm-period coverage will be (a) positively related to initial increases in the

average number of firms a firm mentions in it press releases (by period) but (b)

negatively related to continuing increases.

14

Page 15: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

H10: Firm-period mindshare will be (a) positively related to initial increases in the

average number of firms a firm mentions in it press releases (prior periods) but

(b) negatively related to continuing increases.

H11: Hazard rates of exit will be (a) negatively related to initial increases in the

average number of firms a firm mentions in it press releases (prior periods) but

(b) positively related to continuing increases.

H12: Firm-period sales will be (a) positively related to initial increases in the average

number of firms a firm mentions in it press releases (prior periods) but

(b) negatively related to continuing increases.

DATA AND METHODS

We test our hypotheses in a quantitative study of firms’ attempts to use the press to

position themselves in the early years of the market for computer workstations, a distinct product

market category that emerged in the 1980s. The data we use includes market demographics,

firm-level data, and market- and firm-level measures extracted from analysis of an extensive of

the full text of more than 28,000 news stories and press releases—nearly 60,000 pages of

material—that mention workstations. After briefly explaining the industry context for this study,

we explain the measures and methods used to test our hypotheses.

Site and data: the market for computer workstations, 1980-1990

The computer workstation is a category of computing hardware that developed into a

high-growth niche market starting in the early 1980s. Situated between minicomputers and

personal computers, this niche brought Schumpeterian creative destruction (1934) as its growth

15

Page 16: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

effectively destroyed the value of larger minicomputer manufacturers like Digital Equipment and

Data General while also creating fortunes for the investors and founders of startups like Sun

Microsystems and Apollo Computer, among others. Building on research done at SRI and Xerox

PARC in the 1960s and 1970s, the first commercial workstations of the 1980s packed features of

that era’s minicomputers into a form factor that made it practical to locate the machine in an

individuals workspace and dedicate to the work of a single user whose technically demanding

work could benefit from such power. Yet these products were neither personal computers nor

minicomputers. Compared to contemporary PCs, they were far more powerful, but also

considerably larger and much more expensive. Compared to minicomputers, they were smaller,

cheaper, less powerful, and less appropriate as a shared computing resources.

As it happened, this combination of features was attractive to engineers and scientists

with technical work that benefited from the kind of dedicated computing power workstations

began to offer. As a result, sales grew briskly through the early 1980s, and by the mid- to late -

1980s, the market reached an initial shake-out, though the introduction of lower-end

workstations based on personal computer hardware standards did drive a second round of entry

beginning in the early 1990s. By the mid-late 1990s, the distinctiveness of the workstation

classification was eroded as rapid improvements in personal computer performance blurred the

lines between these two categories and the growth of the Internet fueled demand for servers that

incorporated key features of workstations into packages designed to serve networks rather than

users. For more background on the market for computer workstations, see Bell (1986) for an

explanation of the technology evolution that led to the category, Goldberg (1988) for a

discussion of the history of the market, and Sorenson (2000; 1997) for a study of changing

strategy and organizational form in this market.

16

Page 17: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Period. In keeping with this study’s focus on the role of institutions in the work of

making markets, analysis is limited to the early formation period, from 1980-1990. This choice

offers three benefits. First, it allows us to pick up the first commercial workstations (1981).

Second, it encompasses the entire formation period of this fast-growing market. Third, it avoids

confusion with the blurring of the distinctions between PCs and workstations that occurred in the

early 1990s—itself is an interesting problem for another study.

Demographic Data. Data on workstation entrants were collected from DataSources, an

industry catalog that provided basic data such as entry and exit dates and size (employees).

Demographic data not available from DataSources were developed through additional archival

research, mostly from LexisNexis sources. Gaps in firm-period covariates were also filled this

way when possible, with linear interpolation used to fill gaps in size and sales data when no other

information could be found. From 1980 to 1990, 107 firms entered, and 54 exited, following the

early part of the familiar density dependence pattern (Carroll and Hannan 1989); see Figure 2.

Media Coverage. To develop data about firms’ media coverage and publicity-seeking

efforts, a large storybase was assembled from the news stories and press releases available in

relevant LexisNexis sources. The storybase included 28,292 separate items gather from both

newswires and a variety of publications, including national news magazines, daily newspapers,

mainstream business publications, and more specialized trade publications. Press releases and

news stories were treated differently, and the press releases of market entrants were identified so

their frequency and contents could be analyzed to produce firm-level measures of publicity-

seeking activity. Altogether, the storybase totaled approximately 60,000 pages of single-spaced

text.

17

Page 18: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Methods, Measures and Models

Our study analyzes the impact of three independent variables: the extent of firms’ efforts

to solicit media coverage, the extent of their use of PR agency expertise, and the extent to which

they mention only themselves or others as well in their public information releases. With key

controls, we analyze the relationship between these variables and four dependent variables:

coverage, mindshare, market exit, and sales. We now explain the measures used for these

variables and, where applicable, we briefly describe the methods used to develop them.

Independent variables and controls

Software developed by the first author was used to analyze the storybase to develop firm-

the period measures of publicity-seeking activity and its impact on coverage and positioning.

Since the variables themselves can be explained without going into the details of the tool and its

method, we forego discussion of that in this paper. (For more detail, see Kennedy 2005.)

Releases. To measure firms’ efforts to leverage the press by supplying information to

them, we counted the number of press releases issued by each entrant by calendar quarter; for

brevity, we refer to this variable as releases. To test for the hypothesized curvilinear impact of

supplying information to reporters, we use both releases and its square, releases2. By way of

background, a press release is essentially the company line written up as though it were a news

story. For reporters and firms alike, it is an institutionalized way of supplying information firms

want to publicize in a form that is easy for reporters to use. While our information acknowledged

that press releases are sometimes run as news stories with little to no editing, the consensus

among our information was that this is the exception and generally limited to highly targeted

trade publications that print a lot of product news as a service to their readers.

18

Page 19: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

There are pros and cons to using releases as a proxy for the extent of firms’ efforts to

court media coverage. On the plus side, it provides a very complete record of the occasions

when firms are out selling their stories to the media and the public. Our field data suggest the

issuance of press releases is a ubiquitous element of publicity campaigns that range widely in

scope and complexity. At their simplest, they may involve little more than issuing a press

release. Since more is generally needed to get the scarce attention of reporters’ and their

audiences, firms generally do much more, ranging calling reporters and editors to meeting with

them in-person to sending them elaborate press kits that may include demonstration products and

other promotional materials. Frequently, these activities are planned to coincide with elaborate

events at trade shows, conventions, or special occasions set apart from wider gatherings. On the

negative side, therefore, using a simple count of releases misses the variance created by the

additional activity that can go with releases. While more aggressive publicity campaigns appear

to be correlated with heavier use of press releases, we do not have the data to characterize and

analyze the richness of publicity campaign activity. If we could get that data, we would expect

to explain more variance more precisely; any results we find without it are likely understating the

significance of these richer activities. In any case, we want to be clear that we view releases as a

proxy for this activity. According to our field data, it is not the case that releases alone do the

work of getting reporters’ attention.

PR Usage (Credits). Since a press release is a solicitation for coverage written as a story

containing information that will generally be incorporated into another story, one of its key

features is contact information reporters can use to get to someone charged with giving quick

answers to their questions. Especially for stories written on tight daily deadlines, firms wanting

to support reporters writing about them must have an easy-to-follow procedure for providing

same-day answers to reporter questions—preferably with turn times measured in minutes rather

19

Page 20: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

than hours or day parts. Consequently, firms will often use a PR firm to ensure they are able to

provide the kind of prompt responses that make it easier for reporters to write about them. When

they do, the contact information indicates the name of the PR firm engaged to help. Thus, it is

possible to code each release for whether it is supported by a PR firm or not, and it is also

possible to develop firm-period measures of the PR usage. In effect, this contact information

enables us to track and credit PR firms for the work they do for clients. We count these credits

are use them as a proxy for PR usage. As with releases, we test for the hypothesized curvilinear

impact of PR usage by including both credits and its square, credits2.

As with releases, there are pros and cons to using contact credits as a measure of PR

usage. First, there are times when PR firms are hired to do media relations when they may not

have had much input into the overall campaign. Though this is generally not desirable for PR

firms since media relations work is not a highly leverage activity for the firm, it does happen.

Second, there are times when a PR firm will be engaged for publicity campaign planning but not

for the media relations task that leads to credits on releases. Thus, using credits to PR firms as a

proxy for PR Usage mixes cases that may understate usage with those that where usage may

have entailed little real counsel. With these limitations in mind, we have used credits to PR

firms because it is the best proxy we have for the extent to which firms are turning to the

brokerage capability and inside knowledge skilled PR agents can bring.

Referencing Rivals: Average Release Density (ARD). The features of press releases make

it simple to associate them with their issuers, and this makes it possible to analyze the content of

releases to measure firms’ tendencies to talk only about themselves or also about the

competition. Using the methods and tools mentioned above, we counted the number of firms

mentioned in each press release and computed a firm-period measure of average release density,

the average number of firms a firm mentions in all its press releases for a given period. To 20

Page 21: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

handle problems of limited within-risk set variance that arise in the hazard models, we apply

geometric moving average to ARD, as follows:

∑=

+−=

t

jj

jt

1

)1(2

1

ARDGMA(ARD)

In addition, the smoothed ARD is a more stable indicator of firms’ tendencies to mention only

themselves or themselves plus one or two other firms who have also entered the market. Both

ARD and its square are used to test for the hypothesized curvilinear relationship.

Geometric smoothing of releases and credits. For analyses of coverage, we wanted to

assess the effects of current period release activity on coverage2, so releases and credits are not

smoothed in this analysis. (We ran models for these dependent variables using both smoothed

and unsmoothed ARD and its square; results are reported below.) To analyze how firms’

patterns of releases activity impacted sales and survival, however, we applied a geometric

moving average to releases and credits, just as for ARD (see above).

Controls

Controls for size and experience in the market were obtained from DataSources and

follow-up research; these were updated yearly. Linear interpolation was used to fill in missing

values.

Dependent variables

We used four dependent variables: coverage, a measure of relative attention we call

mindshare, exit events, and sales. Each is discussed briefly.

2 We also used ran the model using lagged coverage; the pattern of results was the same. Field data

strongly suggest using current period coverage is the correct specification, however, because (1) press releases about

products are always issued before the events that drive coverage and (2) rarely kept around for any length of time.21

Page 22: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Coverage. Firm-period coverage was measured by calendar quarter as the number of

stories in which it was mentioned.

Mindshare. Working from the Oxford English Dictionary’s definition of mindshare as

“[C]onsumer awareness of a particular product or brand, esp. compared to the profile enjoyed by

competitors' products,” we measured mindshare by calendar quarter as degrees centrality in a

firm-by-firm network capturing who is co-mentioned with whom. Across all stories in a given

analysis period, therefore, the organizations with the highest mindshare are those most often co-

mentioned with any other workstation producer in all of that period’s stories about workstations.

Since both coverage were positive integer counts with overdispersion (variance greater than the

mean), we estimated results using cross-sectional time-series negative binomial regressions.

Market exit. For survival, we coded exit as 1 for the period in which organizations either

failed outright, went bankrupt, or were sold; exit was 0 otherwise for every period from entry to

either exit or the end of the analysis window.

Sales. The sales data provided in DataSources is, as is the case in financial reporting

more generally, generally not broken out by product market. As a result, sales figures for larger

de alio entrants mostly include sales for firms’ entire product portfolios, making them too noisy

to use reliably. For de novo firms, however, this is not a problem. Therefore, rather than skip

the analysis altogether, we limit analysis to de novo firms. We expect results for this unique set

will not be generalizable to all firms, but we provide the analysis anyway since we think the de

novo sub-sample is theoretically and socially significant. In discussion, we elaborate and note

how and why we expect publicity effects de novo and de alio firms differently.

22

Page 23: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Models and estimation

Since dependent variables for coverage and mindshare are non-negative integer counts,

Poisson regression would be a natural choice for model estimation, but it could not be used

because the data violate its stringent equidispersion assumption (G2=8144.4, p<.01). To avoid

problems of deflated coefficient standard errors and spurious significance (Cameron and Trivedi

1986), negative binomial models were used instead. The data are organized as a longitudinally

unfolding panel of cross-sections, so we used the cross-sectional time series version negative

binomial regression, available in STATA. Furthermore, fixed effects models also control for

firm-specific capabilities that impact product quality. The conventional engineering perspective

on success in high technology markets is that it is all about product leadership. To the extent that

product leadership is a stable quality of firms, using fixed effects estimation provides the

opportunity to test whether approaches to the publicity game contribute to coverage, centrality

and success over and above a company’s base level of product leadership. Therefore, cross-

sectional time-series fixed-effect negative binomial models are used to estimate equations for

coverage and mindshare.

Piecewise exponential hazard rate models are used to estimate equations for the hazard

rate of market exit. These models offer the advantage of allowing researchers to avoid

specifying (or mis-specifying) the functional form of the probability distribution used to predict

‘arrival’ of failure events. While Cox regression (1972) is a popular method for getting around

problems of unknown probability distributions, analysis of Schoenfeld residuals show that the

workstation failure data violate its underlying proportional hazards assumption violated (p < .05).

In piecewise models, the baseline hazard rate is held constant within predefined time pieces

while allowing it to vary freely from piece to piece, so the model makes neither distributional nor

proportionality assumptions. Instead, models estimate aggregate hazard rates in terms of the 23

Page 24: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

contributions made by separate hazard rates observed for predetermined time ‘pieces’ (Blossfeld

and Rohwer 1995; Tuma and Hannan 1984). The method requires that data be organized

according to discrete time pieces, which fits nicely with a quarterly analysis of press release

activity and media coverage; the quarterly analysis periods selected for this study of the

workstation market are small enough to allay concerns about baseline hazard rate change within

periods. Besides handling concerns about misspecification of probability distributions, the

piecewise constant hazard rate model properly controls for duration dependence and right

censoring. Estimation routines for STATA were written by Jesper Sorensen (1999).

RESULTS

Tables 1-4 give descriptive statistics, one table each for the variables in analyses of

coverage, mindshare, exit and sales; Tables 5-8, respectively, give results for each analysis.

Analysis of Coverage (Table 5). Results support the hypothesized curvilinear relationship

between coverage and all three independent variables – releases (H1), credits (H5), and ARD

(H9). As predicted (H1), coverage was curvilinearly related to firms’ efforts to court it, as

measured by firm-period counts of through press releases (p <. 001). Also as predicted (H5),

coverage was also curvilinearly related to firms’ use of PR, as measured by firm-period counts of

press release credits to PR firms, though the predicted negative effects of increasing use of PR

were only marginally significant (p < .10); in contrast, the linear term for credits had greater

statistical significance (p < .01). Again, results also show coverage is curvilinearly related to

ARDgma (p < .05). When adding each variable to the base model with controls only, model

improvement statistics are significant at the (p < .001), (p < .10), and (p < .01) levels, in

succession. Of the three variables, coefficients for PR usage are larger than coefficients for

24

Page 25: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

releases but smaller than coefficients for ARD, but this is misleading since both releases and

credits have much greater variance (min-max range of 0-32 and 0-9, respectively) than ARD

(min-max range of 0-4.285).

Analysis of Mindshare (Table 6). With significant model improvement (χ2 = 6.967, d.f. =

2; p < .05), adding releases and its square to the base model with controls only (column 2) shows

support for the hypothesized (H2) curvilinear relationship between mindshare and firms’ efforts

to court coverage (p < .05, p < .01). In contrast, the hypothesized (H6) curvilinear relationship

between mindshare and PR usage was not supported. As in the analyses of coverage, however,

adding ARD to the model did support the hypothesized (H10) curvilinear relationship between

mindshare and ARD (p < .001), with significant model improvement (χ2 = 21.710, d.f. = 2;

p[<].05).

Analysis of Exit Rate (Table 7). Regardless of lag, we did not expect a single period’s

release activity to have much impact on likelihood of exit, so we use a geometrically smoothed

version of releases and its square in exit analyses; this captures firms’ more or less stable

tendencies to court media coverage. Though results show the predicted (H3) curvilinear

relationship between exit rate and releases (p < .10 for releases, p < .05 for its square), model

improvement statistic was not significant (χ2 = 2.526, d.f. = 2; p = .28). Surprisingly, however,

results do support the hypothesized (H7) relationship curvilinear relationship between exit rate

and PR usage, as measure by firms’ press releases’ credits to PR agents (columns 3 and 4: p

< .05; χ2 = 9.300, d.f. = 2; p[=].01). This effect does not hold up when ARD is added, but adding

ARD and its square to the model with releases and its square (column 2) supports the

25

Page 26: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

hypothesized relationship between exit rate and ARD its square (p < .05), though model

improvement is only marginally significant (χ2 = 5.011, d.f. = 2; p[< .10).

Analysis of Sales – de novo firms only (Table 8). Again, we did not expect a single

period’s release activity or PR use to impact sales, so we estimated sales using geometrically

smoothed versions of the linear and second order terms for variables releases and credits; these

terms capture firms’ more or less stable tendencies to court media coverage and / or use PR.

Results in column 2 show strong support for the predicted (H4) curvilinear relationship between

sales and releases (p < .01), with significant model improvement (χ2 = 38.757, d.f. = 2; p < .001).

Similarly, the results in columns 3 and 4 show strong support for the predicted (H8) curvilinear

relationship between sales and PR usage, as measured by credits to PR agents in press releases

(p[<].01), again with significant model improvement (χ2 = 8.543, d.f. = 2; p < .05). Adding ARD

and its square to the model with releases provides limited support for the predicted (H12)

relationship between sales and ARD: model coefficients are significant (p < .05), but model

improvement is not good (χ2 = 3.525, d.f. = 2; p = .17).

DISCUSSION AND CONCLUSION

In this paper, we have explored the limits of firms’ attempts to leverage institutions to

enact strategic changes like the creation of new markets. Working from field study, we

developed hypotheses relating three different independent variables to four different dependent

variables, allowing us several views of the extent and effectiveness of firms’ attempts to leverage

the institutions of business journalism to their advantage. Overall, the picture that emerges from

these qualitative and quantitative analyses fits together nicely to extend our understanding of

26

Page 27: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

institutions and the role they play in both structuring and enabling the process of naming and

claiming strategic positions—and, of course, realizing the competitive outcomes that come with

these positions.

Limitations and directions for further study. Though the accumulation of qualitative and

quantitative evidence is considerable, it is still ultimately a look at a limited set of factors in the

emergence of a single product market category. Other authors have explained the formation of

technology markets in terms of, for example, the use of “open systems” approaches that facilitate

shared investment in expensive research and development paths (see, for example, Garud 1993).

In keeping with this, it could be that the value of being publicly associated with rivals who

together validate each other’s departure from familiar product market categories pales in

comparison to having the right product or product strategy. We believe there are tradeoffs in

these relationships that allow clear product superiority and clever market formation strategy to

substitute for each other, and our work in this paper does not address these tradeoffs. Generally

speaking, however, we believe it is rare that a lone-wolf innovator breaks from the pack with a

product whose superiority is so easy to evaluate that it is obvious without points of comparison.

As one of our informants put it,

“You can’t create a market without competition. … You need the ‘unlike’ statement.

unlike so-and-so, … You have to have a point of comparison. Without it, you’re dead.

Nobody will pay attention.”

Maybe the needed points of comparison could be substitutes in categories subject to what

Schumpeter (1934) called creative destruction as the new thing grows, even if growth comes

without the sort of direct competition generally regarded as entry into emerging markets. At the

same time, maybe growth is limited without that entry, even though that brings competition.

27

Page 28: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Certainly, this is the implication of density dependence findings in organization ecology (for a

summary, see Carroll and Hannan 2000). To sort out these questions, further study will have to

move the level of analysis from the firm up to markets, including both successes and failures.

Contributions. Where previous research has suggested firms are able to reconfigure

conflicting aspects of institutions in flux during relatively rare times of upheaval, we have shown

that contradictions exist even in stable institutions and that they provide, at least in the case of

business journalism, opportunities to leverage the institution for competitive gains even as they

also barriers to attempts to over-use those leverage opportunities. While institutionalized norms

for business journalism stipulate the avoidance of explicit bargains with sources, the practical

workaday world of journalists requires numerous small departures from this ideal. As reporters

well know, there is no such thing as a free lunch. Indeed, the sources on whom they rely for

freely volunteered information in fact charge a subtle price for the information they provide: it is

reliably slanted toward the source. As long as reporters cooperate with their sources, they will

tend to pass this slanted information on to their audiences. Thus, unless and until it pays reporters

to betray their sources by doing negative stories on them, the news—at least the business news—

will tend to relay a source-slanted bias we refer to as the source costs of the information

exchange system sketched in Figure 1. The complex of information flows in that diamond only

works when reporters and sources are able to find a cooperative groove in which they each give

the other what they want, even if only for a time.

In our view, this is a theoretically and socially important example of an institution

providing opportunities for leverage even as it lays down limits to that leverage that tend to

perpetuate the institution. Thus, it may be that the paradox of institutions—that they provide both

limits to action and levers for it—is exactly what supports their durability.

28

Page 29: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

FIGURES AND TABLES

Figure 1

stories, claims

feedback, inquiries

Reporters

PR Agents

AudiencesFirms

Information flows in market analysis:anatomy of an economy of reciprocity

stories, claims

feedback, inquiries

Reporters

PR Agents

AudiencesFirms

Information flows in market analysis:anatomy of an economy of reciprocity

29

Page 30: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Figure 2

30

Page 31: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Table 1 CORRELATIONS VARIABLE OBS. MEAN S.D. MIN MAX (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) 1. coverage 1251 19.741 60.279 0 590 2. Coveraget-1 1251 18.890 58.388 0 590 .977 3. ln(Size) 1251 7.174 2.468 2.197 12.913 .443 .440 4. ln(Age at Entry) 1251 7.983 1.421 3.871 10.486 .202 .201 .615 5. Experience (qtrs) 1251 12.551 8.027 1 37 .089 .097 .184 .011 6. De novo (=1) 1251 0.345 0.476 0 1 -.095 -.094 -.471 -.738 .108 7. Releases (No PR) 1251 1.515 3.356 0 32 .610 .599 .371 .090 .257 .045 8. Releases2 (No PR) 1251 13.547 59.630 0 1024 .538 .532 .230 .020 .188 .069 .878 9. Credits 1251 0.230 0.736 0 9 .068 .061 .028 -.129 .109 .159 .161 .096 10. Credits2 1251 0.595 3.739 0 81 .016 .011 -.006 -.101 .069 .124 .099 .048 .852 11. ARDgma 1251 0.675 0.740 0 4.285 .238 .236 .363 .101 .248 -.010 .511 .267 .321 .183 12. ARD2

gma 1251 1.003 1.673 0 18.363 .145 .147 .218 .013 .182 .017 .402 .206 .258 .168 .910

Table 2 CORRELATIONS VARIABLE OBS MEAN S.D. MIN MAX (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

1. Mindshare 1246 9.704 10.987 0 50

2. Mindsharet-1 1246 9.194 10.942 0 50 .873 3. ln(Size) 1246 7.278 2.426 2.197 12.913 .563 .549 4. ln(Age at Entry) 1246 8.043 1.406 3.871 10.486 .211 .202 .617 5. Experience (qrs) 1246 12.210 8.141 0 37 .269 .301 .149 -.015 6. Releases (No PR) 1246 1.430 2.949 0 21.547 .572 .551 .367 .080 .261 7. Releases2 (No PR) 1246 10.731 43.580 0 464.252 .427 .424 .228 .013 .193 .878 8. Creditsgma 1246 0.212 0.513 0 4.953 .140 .127 .032 -.131 .115 .158 .097 9. Cedits2

gma 1246 0.307 1.416 0 24.533 .059 .053 .001 -.103 .067 .085 .044 .864 10. ARDgma 1246 0.670 0.747 0 4.285 .406 .380 .343 .081 .230 .509 .265 .314 .165 11. ARD2

gma 1246 1.007 1.689 0 18.363 .274 .255 .201 .006 .154 .395 .201 .236 .119 .911

31

Page 32: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Table 3 CORRELATIONS VARIABLE OBS MEAN S.D. MIN MAX (1) (2) (3) (4) (5) (6) (7) (8) (9)

1. Exit (=1) 1246 0.001 0.028 0 1 2. in(Age at Entry) 1246 8.043 1.406 3.871 10.486 -.047 3. Size 1246 7.278 2.426 2.197 12.913 -.036 .617 4. Coveraget-1 1207 19.331 59.318 0 590 -.008 .195 .440 5. Releasesgma 1246 1.641 3.105 0 22.522 -.015 .059 .384 .628 6. Releases2gma 1246 12.326 48.152 0 507.245 -.007 -.007 .245 .590 .896 7. Releasesgma (No PR) 1246 1.430 2.949 0 21.547 -.014 .092 .398 .645 .987 .897 8. Releases2gma (No PR) 1246 10.731 43.580 0 464.252 -.007 .013 .255 .605 .886 .994 .900 9. ARDgma 1246 0.670 0.747 0 4.285 -.025 .081 .343 .238 .569 .302 .535 .290 10. ARD2

gma 1246 1.007 1.689 0 18.363 -.017 .006 .201 .146 .442 .231 .417 .222 .911

Table 4 CORRELATIONS VARIABLE OBS. MEAN S.D. MIN MAX (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)

1. Sales 230 124.461 185.642 2 1052

2. (Sales / 1,000)t-1 230 0.117 0.174 0.001 1.052 .972 3. ln(Coveraget-1) 230 -2.678 6.476 -11.513 5.389 .491 .487 4. ln(Size) 230 5.991 1.439 3.332 9.012 .757 .748 .591 5. ln(Age at Entry) 230 780.230 286.568 181 1552 -.217 -.215 -.251 -.200 6. Experience (qtrs) 230 15.887 8.371 1 32 .559 .573 .534 .701 -.243 7. Releasesgma 230 2.916 4.852 0 22.522 .836 .812 .434 .673 -.254 .425 8. Releases2gma 230 31.944 90.889 0 507.245 .811 .778 .352 .549 -.165 .323 .935 9. Releasesgma (No PR) 230 2.422 4.551 0 21.068 .830 .803 .449 .662 -.248 .415 .984 .940 10. Releases2gma (No PR) 230 26.490 79.850 0 443.848 .799 .763 .356 .540 -.160 .319 .919 .994 .938 11. Creditsgma 230 0.493 0.885 0 4.953 .315 .325 .069 .286 -.115 .196 .421 .296 .254 .218 12. Credits2

gma 230 1.023 3.004 0 24.533 .327 .338 .135 .295 -.081 .253 .373 .300 .278 .266 .617 13. ARDgma 230 0.801 0.845 0 3.995 .480 .476 .385 .580 -.319 .533 .600 .396 .573 .381 .340 .303 14. ARD2

gma 230 1.353 2.174 0 15.957 .337 .331 .281 .379 -.311 .401 .448 .293 .439 .282 .196 .127 .913

32

Page 33: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Table 5

Page 34: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Table 6

34

Page 35: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Table 7

35

Page 36: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Table 8

Analysis of Sales (for de novo firms) † Market for Computer Workstations, 1980-1990

(1) (2) (3) (4) (5) Controls Rel. Rel~PR PR ARD (Sales / 1,000)t-1 0.058 -0.015 0.021 -0.021 0.076 (0.093) (0.102) (0.107) (0.109) (0.096) ln(coveraget-1) -0.003 -0.004 -0.004 -0.003 -0.002 (0.005) (0.005) (0.005) (0.005) (0.004) Size 0.726 0.706 0.732 0.694 0.670 (0.051)** (0.037)** (0.042)** (0.041)** (0.034)** Age at Entry 0.003 0.004 0.004 0.004 0.004 (0.001)* (0.001)** (0.001)** (0.001)** (0.001)** Experience -0.280 -0.142 -0.177 -0.167 -0.395 (0.172) (0.068)* (0.067)** (0.072)* (0.164)* All Releasesgma 0.088 0.060 (0.013)** (0.018)** (All Releasesgma)2 -0.003 -0.002 (0.000)** (0.001)** Releases No PRgma 0.053 0.046 (0.017)** (0.017)** (Releases No PRgma)2 -0.002 -0.002 (0.001)** (0.001)** Creditsgma 0.148 (0.067)* Creditsgma

2 -0.025 (0.018) ARDgma 0.212 (0.082)** ARDgma

2 -0.070 (0.029)* Constant -3.282 -4.046 -4.138 -3.987 -2.930 (1.543)* (0.829)** (0.957)** (0.923)** (0.923)** Observations 230 230 230 230 230 Firms 20 20 20 20 20 Periods 41 41 41 41 41 Obs. 230 230 230 230 230 d.f. 38 40 40 42 42 LR chi2 38.757 9.817 8.543 3.525 Prob > Chi2 0.000*** 0.007** 0.014* 0.172

† Sales as thousands of dollars; cross-sectional time series negative binomial

regression using firm-fixed effects. Coefficients for period dummies omitted.

(a) Variables with the subscript gma are geometric moving averages.

Standard errors in parentheses: + significant at .10; * significant at .05; ** significant at .01; *** significant at .001

36

Page 37: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

REFERENCES

Baum, Joel A. C. 1999. "The Rise of Chain Nursing Homes in Ontario, 1971-1996." Social Forces 78:543-583.

Baum, Joel A. C. and Christine Oliver. 1992. "Institutional Embeddedness and the Dynamics of Organizational Populations." American Sociological Review 57:540-559.

Baum, Joel A.C. and Christine Oliver. 1991. "Institutional Linkages and Organizational Mortality." Administrative Science Quarterly 36:187-218.

Bell, G. Gordon. 1986. "Toward a History of Personal Workstations." Pp. 1-17 in Proceedings of the ACM Conference on the History of Personal Workstations (Palo Alto, CA). NY: ACM Press.

Blossfeld, Hans-Peter and Götz Rohwer. 1995. Techniques of Event History Modeling. Mahwah, NJ: Lawrence Erlbaum.

Cameron, A.C. and P.K. Trivedi. 1986. "Econometric models based on count data: comparisons and applications of some estimators and tests." Journal of Applied Econometrics 1:29-53.

Carpenter, Gregory S. and Kent Nakamoto. 1989. "Consumer Preference Formation and Pioneering Advantage." Journal of Marketing Research 26:285-298.

Carroll, Glenn and Michael T. Hannan. 2000. The demography of corporations and industries. Princeton, N.J.: Princeton University Press.

Carroll, Glenn R. and Michael T. Hannan. 1989. "Density Dependence in the Evolution of Populations of Newspaper Organizations." American Sociological Review 54:524-541.

Clemens, Elisabeth S. and James M. Cook. 1999. "Politics and Institutionalism: Explaining Durability and Change." Annual Review of Sociology 25:26.

Clemens, Elisabeth Stephanie. 1997. The people's lobby: organizational innovation and the rise of interest group politics in the United States, 1890-1925. Chicago: University of Chicago Press.

Cox, D. R. 1972. "Regression Models and Life Tables." Journal of the Royal Statistical Society Ser. B, 34:187-220.

DiMaggio, Paul and Walter W. Powell. 1983. "The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields." American Sociological Review 48:147-160.

Friedland, Roger and Robert R. Alford. 1991. "Bringing society back in: symbols, practices, and institutional contradictions." Pp. 232-263 in The New Institutionalism in Organizational Analysis, edited by W. W. Powell and P. J. DiMaggio. Chicago: University of Chicago Press.

37

Page 38: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Garud, Raghu. 1993. "Changing Competitive Dynamics in Network Industries: an exploration of Sun Microsystems' open systems strategy." Strategic Management Journal 14:351-369.

Goldberg, Adele. 1988. A History of Personal Workstations. NY: ACM Press.

Kennedy, Mark Thomas. 2005. "Behind the One-Way Mirror: Refraction in the construction of product market categories." Poetics 33:201-226.

Lieberman, Marvin B. and David B. Montgomery. 1998. "First-mover (dis)advantages: Retrospective and link with the resource-based view." Strategic Management Journal 19:1111-1125.

Meyer, John W. and Brian Rowan. 1977. "Institutionalized Organizations: Formal Structure as Myth and Ceremony." American Sociological Review 83:340-363.

Meyer, John W. and W. Richard Scott. 1983. Organizational environments: ritual and rationality. Beverly Hills: Sage.

Polanyi, Karl. [1944] 1985. The great transformation. Boston: Beacon Press.

Porac, Joseph F. and Jose Antonio Rosa. 1996. "Rivalry, Industry Models, and the Cognitive Embeddedness of the Comparable Firm." Pp. 363-88. in Advances in Strategic Management, vol. 13, edited by J. A. C. Baum and J. E. Dutton. Greenwich, CT.: JAI Press.

Porac, Joseph F. and Howard Thomas. 1990. "Taxonomic Mental Models in Competitor Definition." Academy of Management Review 15:224-240.

Porac, Joseph F., Howard Thomas, Fiona Wilson, Douglas Paton, and Alaina Kanfer. 1995. "Rivalry and the industry model of Scottish knitwear producers." Administrative Science Quarterly 40:203-227.

Rosa, Jose Antonio, Joseph F. Porac, Jelena Runser-Spanjol, and Michael S. Saxon. 1999. "Sociocognitive dynamics in a product market." Journal of Marketing 63:64-77.

Schumpeter, Joseph A. 1934. The Theory of Economic Development. Cambridge, MA: Harvard University Press.

Scott, W. Richard. 1995. Institutions and organizations. Thousand Oaks: Sage.

Sorenson, Jesper. 1999. "STPIECE: Stata module to estimate piecewise-constant hazard rate models." Boston: Statistical Software Components (Boston College Department of Economics).

Sorenson, Olav. 2000. "Letting the market work for you: an evolutionary perspective on product strategy." Strategic Management Journal 21:577-592.

38

Page 39: Leveraging Stable Institutions for Strategic Change:€¦  · Web viewSituated between minicomputers and personal computers, this niche brought Schumpeterian creative destruction

Sorenson, Olav Johann. 1997. "The Complexity Catastrophe in the Computer Industry: Interdependence and Adaptability in Organizational Evolution." Ph.D. Thesis, Sociology, Stanford University, Stanford, CA.

Tuma, Nancy and Michael T. Hannan. 1984. Social Dynamics: Models and Methods. NY: Academic Press.

Westphal, James D. and Edward J. Zajac. 1994. "Substance and symbolism in CEOs' long-term incentive plans." Administrative Science Quarterly 39:367-90.

White, Harrison C. 1981. "Where do markets come from?" American Journal of Sociology 87:517-547.

—. 2002. Markets from Networks: Socioeconomic models of production. Princeton: Princeton University press.

Zajac, Edward J. and James Westphal. 1998. "Symbolic Management of Stockholders: Corporate Governance Reforms and Shareholder Reactions." Administrative Science Quarterly 43:127-153.

Zajac, Edward J. and James D. Westphal. 1995. "Accounting for the explanations of CEO compensation: substance and symbolism." Administrative Science Quarterly 40:283-308.

Zuckerman, Ezra W. 1999. "The categorical imperative: securities analysts and the illegitimacy discount." American Journal of Sociology 104:1398-438.

39