Academy of Management

Nonmarket Strategy Performance: Evidence from U.S.
Electric Utilities
Author(s): Jean-Philippe Bonardi, Guy L. F. Holburn and Richard G. Vanden Bergh
Source: The Academy of Management Journal, Vol. 49, No. 6 (Dec., 2006), pp. 1209-1228
Published by: Academy of Management
Stable URL: http://www.jstor.org/stable/20159828 .
Accessed: 07/04/2014 00:59
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Academy of Management is collaborating with JSTOR to digitize, preserve and extend access to The Academy
of Management Journal.
http://www.jstor.org
This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

All use subject to JSTOR Terms and Conditions
Academy ofManagement Journal
2006, Vol. 49, No. 6, 1209-1228.
NONMARKET STRATEGY PERFORMANCE:

EVIDENCE FROM U.S. ELECTRIC UTILITIES
JEAN-PHILIPPEBONARDI
GUY L. F. HOLBURN
University ofWestern Ontario
RICHARD G. VANDEN BERGH

University of Vermont
This paper develops and tests a theory of the performance determinants of a firm's
nonmarket strategy in shaping public policy outcomes. Building on the concept of
political market attractiveness, we argue that nonmarket performance is influenced by
both the characteristics of a firm's regulatory and political environment, especially
rivalry among interest groups or politicians, and by internal capabilities that enable a
firm tomitigate political transaction costs. Using data on regulatory filings for rate
increases made by U.S. electric utilities over a 13-year period, we find empirical
support for our approach.
Although the last decade has witnessed in & Rehbein, 1991; Schuler, 1996; Schuler, Rehbein,
creased interest in the design and implementation & Cramer, 2002). Despite these studies, however,
of firms' nonmarket strategies?defined as the co little analysis has directly examined the determi
ordinated actions firms undertake in public policy nants of actual performance (Keim & Baysinger,
arenas (Baron, 2003; Baysinger, 1984; Hillman, 1988). As Getz noted in a survey of extant work in
Sch?ler, & Keim, 2004; Shaffer, 1995)?extant re this stream, "If political action is ever to be fully
search has remained relatively silent regarding the
integrated with strategic planning and organization
actual performance of such strategies. By perfor al behavior or practically), much
(intellectually
mance, we mean the ability of firms to effect favor more empirical work on effectiveness will need to
able public policy decisions. For instance, firms be done" (1997: 64). Our objective in this paper is
may seek legislative or regulatory support for spe thus to extend the current literature by exploring,
cific environmental emissions standards, import both and nonmarket
theoretically empirically,
tariff policies, antitrust decisions, or regulated
strategy performance.
rates. Relative to a given status quo policy, perfor
A natural question iswhy the academic literature
mance measures the ability of a firm to either
has produced so little investigation into the issue of
achieve policy closer to its preferred position or
performance. After all, this is a critical managerial
block proposals thatmove policy further from that
issue. We advance two explanations. At a theoret
position. Considerable attention has been paid to
ical level, the field has until recently lacked a uni
firms' decisions regarding investments in, and the
fying conceptual framework that analyzes the de
structure of, nonmarket strategies (Bonardi, 2004; terminants of nonmarket strategy performance
de Figueiredo & Tiller, 2001; Grier, Munger, & Rob
(Lord, 2000). Recent scholarship, however, has pro
erts, 1994; Hillman & Hitt, 1999; Holburn & Vanden
posed a framework of "political markets" in which
Bergh, 2002, 2004; Keim & Zeithaml, 1986; Lenway
interactions of demanders (firms, consumers,
unions, activists, etc.) and suppliers (elected poli
We thank participants at theAtlanta Competitive Ad ticians, regulatory agents, and courts) shape public
vantage Conference, the Academy ofManagement annual
policies (Bonardi, Hillman, & Keim, 2005). In this
meeting, and the International Society for theNew Insti view, suppliers such as legislators implicitly trade
tutional Economics forhelpful comments on the paper. votes on legislative bills for electorally valuable
Adam Fremeth provided exceptional research assistance. resources such as campaign contributions. In a
An earlier version was published in the proceedings of
spirit similar to Porter's structural industry analy
the 2006 annual meeting of theAcademy ofManagement.
We are grateful to the editor and the three anonymous sis (Porter, 1980), the framework assesses the inher
reviewers for their feedback and guidance. The research ent attractiveness
of operating in different types of
was supported by a grant from the Social Sciences and political markets.
Structural characteristics, such
Humanities Research Council of Canada. as rivalry among demanders or suppliers, make
1209

1210 Academy of Management Journal December
political markets more or less attractive from a THEORETICAL BACKGROUND

firm's perspective, thereby influencing the firm's In seeking understanding of the factors that drive
decision to engage in nonmarket strategies. nonmarket strategy performance, we find it helpful
Here, we use and build on the political markets
to draw an analogy with the competitive strategy
framework to develop theoretically grounded pre literature. Scholars have argued that a firm's per
dictions regarding the performance of firm nonmar
formance is correlated either with industry attrac
ket strategy. In particular, we extend the framework
tiveness (McGahan & Porter, 1997; Porter, 1980) or
to incorporate other institutional suppliers of pub
with the firm's distinctive capabilities (Barney,
lic policies: regulatory agencies, which have re
1991; Rumelt, 1991; Wernerfelt, 1984). Here, we
sponsibility for designing and implementing poli a similar distinction
draw between external and
cies (Weidenbaum, 2003). Since
agency objectives internal drivers of performance, which we now dis
are not necessarily aligned with those of elected cuss in turn.
politicians, firms may need to adapt their nonmar
ket strategies when interacting mainly with regula
tory agencies. Political Markets and Firms'
We expand the political markets approach also Nonmarket Performance
by exploring how firm-specific capabilities affect
Researchers in economics and political science
nonmarket performance. Several authors, building
have argued that a firm's political environment can
on the resource-based view, have suggested that
be characterized as a marketplace in which de
nonmarket capabilities that draw on firms' internal
manders and suppliers transact over public poli
processes, resources, and knowledge related to po
cies. Originally developed in the 1960s, the politi
litical activities are unevenly distributed among
cal markets approach challenged an axiom
firms and that firms with such nonmarket capabil
common in the economics literature that govern
ities should be more effective in influencing public
ment institutions adopt and implement public pol
policies (Baron, 2003; Dean & Brown, 1995; Hill
icies in the "public interest" (Buchanan & Tullock,
man et al., 2004; Keim & Baysinger, 1988). Here, we
1962; Stigler, 1971). Instead, politicians exchange
build on this general proposition in the context of
policy favors for resources from organized interest
the political markets framework and develop spe
to groups tomaximize their electoral prospects. Valu
cific hypotheses relating nonmarket capabilities able resources include votes from supporting inter
performance. est groups, financial resources, and information, all
The second reason for the paucity of work on
ofwhich can influence election outcomes (Mueller,
nonmarket strategy performance, we speculate, 2003). Since most voters remain rationally ignorant
stems from the difficulty of obtaining data on both
about policy details because of the costs of becom
the structure of firms' nonmarket strategies and
ing fully informed, politicians have some scope to
their performance impact on a particular policy trade policies that deviate from the "public inter
issue. Existing studies have investigated the impact
est" (Aranson, 1990). The implication is that firms,
of nonmarket activities using highly aggregated the of non
through appropriate implementation
measures of firms' performance, such as corporate market strategies, can influence policy makers' de
financial profitability (Hillman, Zardkoohi, & Bier cisions. Figure 1 represents a political market in
man, 1999; Shaffer, Quasney, & Grimm, 2000). a focal firm that wishes to influence a
volving
Here, we overcome the resulting identification and particular public policy.
measurement challenges by using rich information The objectives of the suppliers?such as election
on U.S. electric utilities' nonmarket stategies. We for politicians in democracies?shape the types of
constructed a panel data set that includes specific resources that are valuable in a political market
measures of the performance of a firm's nonmarket & Keim, 1995). In democratic insti
place (Hillman
strategy?in this case, regulatory agency decisions tutional systems, for example, value
politicians
on the financial rate of return that a U.S. electric votes and the resources that generate votes. De
utility may earn?and a precise identification of a manders, including firms, who can provide these
firm's decision to implement a nonmarket strategy: resources have an opportunity to gain favorable
a utility's decision to file a formal request with the Firms design nonmarket strate
policy decisions.
regulatory agency to change its rates. Using this gies, either individually or in concert with other
novel data set, we find support for the validity of firms or groups, to effectively participate in politi
the political markets framework, including firm cal markets, providing votes through, for instance,
specific capabilities, in determining the perfor constituency building; financial support, such as
mance of firms' nonmarket strategies. campaign contributions; and information regarding

2006 Bonardi, Holburn, and Vanden Bergh 1211
FIGURE 1
Demand and Supply Components of a Political Market
Demanders of public policy Suppliers of public policy
Focal Firm Elected Polticians

Nonmarket Strategy\
Votes
Finances
Information
Organized Interest
Groups
Regulatory
e.g., consumers, Agencies
activists, unions
Public Policy
Other Industry Courts
Firms, Entrants
policy consequences and alternatives (Hillman & tutions. From a firm's perspective, then, designing
Hitt, 1999). Bonardi, Hillman, and Keim (2005) ex nonmarket strategies to interface with regulatory
amined the conditions under which the demand agencies presents different challenges from those
and supply sides of the political market for a spe posed by strategies targeted at elected politicians
cific issue are attractive from a firm's perspective. (Baron, 2001). One contribution of this article,
This approach provides an overall framework therefore, will be to better integrate agencies into
within which to study the factors that drive the the political markets framework and to develop
performance of firms' nonmarket strategies. hypotheses regarding how agencies affect the per
formance of a firm's nonmarket strategy.
Political Markets and the Role of

Regulatory Agencies Political Markets and Firms'
Nonmarket Capabilities
Although the political markets approach has
spurred research in both the economics and strate Another limitation of the political markets frame
gic management literatures (Bonardi et al., 2005; de work as developed to date relates to the existence of
Figueiredo & Silverman, 2006), a shortcoming to distinctive nonmarket capabilities within firms.
date has been the relative neglect of the supply side Many studies, especially empirical studies, have
and especially of the role played by regulatory excluded nonmarket capabilities, implicitly treat
agencies. Much of the existing literature examines ing firms as homogeneous entities (Hillman et al.,
how firms or organized interest groups design cam 2004). Following the resource-based view of the
paign contribution strategies aimed at securing the firm, however, several researchers have argued that
support of elected legislators in the legislative pro an important component of nonmarket strategies
cess for proposing, modifying, or vetoing legislative and of their performance lies in firms' internal non
bills. In many industries, however, expert agencies market capabilities (Baron, 2003; Dean & Brown,
have primary responsibility for designing and im 1995; Hillman & Hitt, 1999; Vietor, 1994). Nonmar
plementing public policies through administrative ket capabilities consist of tacit and nontacit knowl
regulations. They are also prohibited from accept edge and skills that enable the firms tomanage the
ing financial resources from the firms they regulate. public policy process and to achieve favorable leg
Furthermore, agency motivations are typically not islative, executive, administrative, and judicial pol
dictated by the ballot box, since executives or leg icy
outcomes.
islatures usually appoint agency heads. Such fac The significance of nonmarket capabilities as a
tors suggest that agency decisions can have impor determinant of nonmarket performance has also
tant consequences formany firms and that agencies not been clearly articulated. Here, we argue that
may behave insti nonmarket are particularly
differently from elected political capabilities important

1212 Academy ofManagement Journal December
in explaining heterogeneity among firms' nonmar their rivals' average wage was much closer to the
ket performance since political markets exhibit extant minimum wage. This asymmetry generated
high transaction costs, or factors that impede the significant lobbying pressures by rivals against
specification, monitoring, or enforcement of trans Wal-Mart's
proposal.1
actions (Dixit, 1996). As North commented, "Polit Second, consumers of a firm's products or ser
ical markets are characterized by imperfect infor vices can also pose a threat by demanding public
mation, subjective models and high transaction policies that lower rates or increase costs through
costs. . . .The political market has been, and con new quality, environmental, safety, or other stan
tinues to be, one in which the actors have an im dards. Although the costs of collective action are
perfect understanding of the issues

affecting them typically substantial for individual consumers,
and equally in which the high costs of transacting consumers that are sufficiently large or that can
prevent the achievement of efficient solutions" obtain public funds can organize against the firm in
(1990: 357). Given the prohibition on explicit con policy arenas (Holburn & Vanden Bergh, 2006; Ol
tracts between special interest groups and politi son, 1965). A third type of nonmarket competition
cians?and hence on judicial enforcement?the stems from interest groups such as unions and en
risks of opportunism and market failure are high vironmental activists. Recent scholarship has
(Dixit, 1996: 53). It is in this context that nonmarket shown that these actors can be tough opponents for
capabilities play
a key role. Firms that develop the firms since a common strategy has been tomobilize
to sustain "trade" in markets, espe the media, a practice that, by providing new infor
ability political
cially by developing credible reputations (Eggert mation to otherwise uninformed voters, enables al
son, 1993), can overcome these intrinsic barriers teration of public perceptions of policy issues (Bo
and successfully implement nonmarket strategies. nardi & Keim, 2005). By making issues more
politically salient, these interest groups can exert
on and appointed bu
HYPOTHESIS DEVELOPMENT powerful pressure politicians
reaucrats. Again, Wal-Mart is an interesting exam
We now build on the political markets frame ple, as many activists and unions have, sometimes
work presented above to develop hypotheses on the communities to oppose or
successfully, pushed
determinants of a firm's nonmarket strategy perfor stall the opening ofWal-Mart superstores through
mance. We focus our arguments around four key out the United States.
factors: the degree of rivalry among demanders, the As opposed interest groups compete more vigor
degree of rivalry among politicians, the resource ously against a firm for their preferred policies,
base of a focal regulatory authority, and finally, the policy makers' bargaining positions improve, en
nonmarket capabilities of a focal firm. them to demand more in return for policy
abling
favors?greater electoral campaign contributions or
grassroots mobilization, for example (Keim &
Demand Side: Interest Groups
Baysinger, 1988). The performance or effectiveness
Mueller argued that "politics in themodern dem of firms' nonmarket expenditures in achieving fa
ocratic state is not a confrontation between two vorable policy outcomes will thus be reduced in
economic classes, but rather a struggle such environments. This leads to our first
polarized
among a plethora of groups with divergent inter hypothesis:
ests" (2003: 472). Firms, in developing nonmarket
Hypothesis 1. Rivalry from competing interest
strategies, may face opposition from several types
groups is negatively related to the performance
of demanders in themarket for public policies (Ma
of a firm's nonmarket strategy.
hon, 1993). First, other firms, either market rivals
or others within the broader structure of a focal
firm's industry, may be disadvantaged, either abso Supply Side: Elected Politicians
lutely or relatively, by the regulatory changes pro
Competitive rivalry for public policies exists
the focal firm. firms that are
posed by Competing not only on the demand side but also on the
either individually or in in
politically organized Recent research
associations can generate high levels of ri supply side of the marketplace.
dustry that elected are more
suggests politicians recep
valry (Stigler, 1971). As an example, during 2005 tive to interest group demands for regulatory fa
Wal-Mart proposed an increase to the minimum
vors when electoral competition or rivalry is
wage. This policy change would have asymmetri
cally affected rivals in the retail industry, as Wal
1
Mart's average wage paid to employees was signif See "Trouble in Wal-Mart's America," the Washing
icantly higher than the current minimum wage, but

ton Post, October 26, 2005.

2006 Bonardi, Holburn, and Vanden 1213
Bergh
strong (Ansolabehere, de Figueiredo, & Snyder, tures or executives, are thus often the central
2003). Greater rivalry between electoral candi point of contact for responding to the require
dates or party coalitions makes candidates more ments of public policies that regulate their busi
willing to "trade" policy favors for campaign con nesses (Holburn & Vanden Bergh, 2004).
tributions or other valuable support that maxi We argue here that the environmental conditions
mizes their chances of election
(Baron, 2001). that enable firms to successfully gain the support of
U.S. steel producers, for instance, substantially regulatory agencies are quite different from those in
increased their lobbying of the Bush presidential the legislative and executive arenas. The distinc
administration in 2002 in search of a tariff on tion stems from the different incentives and con
steel imports. One factor that strengthened their straints that the two types of institutions operate
bargaining power was the existence of extremely within. Regulators are typically appointed rather
tight competition between Republican and Dem than elected, so they do not face the election con
ocrat candidates for House seats in several states straint that can motivate elected politicians' behav
with steel or related industries. Ultimately, in the iors. Existing research suggests that regulators' ob
months before the election, Bush implemented a jective functions are especially multidimensional:
30 percent tariff.2 Naturally, this willingness to regulators may try tomaximize the budgets of their
trade policy favors is conditioned by the broader offices (Niskanen, 1971), expand the numbers of
public saliency of the relevant policies: trading personnel employed, or enhance their career pros
policy with organized interests can come at the pects or
political reputations (Mueller, 2003; Nis
expense of electoral votes if an issue is of partic kanen, 1971; Weatherby, 1971; Weber, 1947). Since
ular concern to voters (Bonardi & Keim, 2005). achieving these objectives depends on the legiti
For less salient policies, firms may press for pol macy that regulators hold within the institutional
icy support in the form of new legislation (or a
system, meta-objective of regulators is to preserve
amendments to existing chamber bills) or in the or increase their legitimacy (Majone, 1996). To do
oversight of regulatory agencies. so, regulators adhere to the procedural constraints

On the other hand, when a political party or that govern their decision making and that are de
coalition has a powerful hold on office through a signed to ensure that regulators implement policies
large and sustained electoral majority?and in accordance with the broad wishes of the en
hence has little need
for additional support from abling legislators (McCubbins & Schwartz, 1984;
special interests?firms are less able to sway pol Weingast & Moran, 1983). Procedural requirements
icy outcomes from an incumbent party's pre relate to the informational basis of regulatory deci
ferred position. Firms, which tend to be better sions: agencies generally must obtain information
politically organized than other interest groups from affected parties, base their final decisions on
(Stigler, 1971), will thus experience a more favor the evidence presented, and publicly announce,
able policy environment when political rivalry along with their rationales, proposed policy
is high. changes (McCubbins, Noll, & Weingast, 1987,
2. Rivalry between politicians 1989).

Hypothesis is
Although such informational requirements en
positively related to the performance of a able legislative committees and executives tomon
firm's nonmarket strategy. itor agency behavior and to prevent arbitrary deci
sions, they also create a resource dependency
Supply Side: Regulatory Agencies relationship between an agency and the firms it
regulates (Pfeffer, 1981, 1992; Pfeffer & Salancik,
Although elected politicians decide the broad
characteristics of public policies,
1978). In particular, regulators depend on firms and
specific details, other interested parties to provide valuable infor
day-to-day implementation, monitoring, and en
mation during regulatory hearings (Mueller, 2003).
forcement activities are delegated to regulatory
A regulatory agency uses this information as evi
agencies in most
jurisdictions. Specifying and
dence in support of its proposals. Without substan
implementing detailed policies typically requires tiation of its policy rulings, an agency risks the
a high degree of information, and agencies are
one organizational mechanism for developing rulings being overturned by the courts, generating
an important loss of legitimacy. The European
sustained policy expertise. From a firm's per
Commission, for example, suffered such a loss in
spective, regulatory agencies, rather than legisla
2002, when three of its decisions against the merg
ers of private companies were voided by the Euro
2
See The Politics of Steel, BBC News, http://news. pean Court of Justice. The Court found that the
bbc.co.uk, March 6, 2002. economic analyses of the mergers' anticompetitive

effects were based on insufficient evidence. These ties that improve their performance in these types
decisions questioned the authority of the Commis of environments (Dean & Brown, 1995). Direct ex
sion and of its head, Mario Monti, and led to its periences with and regulators enable
politicians
reform in 2003.3 Agencies with larger budgets and firms to better understand the patterns of behavior
greater expertise are thus better positioned to inde and preferences of policy makers (Holburn &
pendently obtain their own information, assess Vanden Bergh, 2002; Ring, Lenway, & Govekar,
firms' arguments, and counter firms' policy propos
1990). Some of these capabilities become embed
als (Oliver, 1991). Less well endowed agencies, on ded within managers and employees who are able
the other hand, will be more on the
dependent to leverage their individual experiences. Others be
information provided by firms in formulating their come embedded within firms' operating routines;
decisions, lending a natural bias toward the firms. firms establish codified and uncodified practices
It follows that the greater a regulatory authority's
that reflect prior managerial approaches to resolv
resources, the less dependent it is on firms and the
ing these issues (Boddewyn & Brewer, 1994; Keim
more difficult or costly it becomes for firms to
& Baysinger, 1988). Such capabilities enable firms
obtain favorable agency decisions. Hence:
to alleviate transaction costs and to more effec
Hypothesis 3. The resource base of a relevant tively implement nonmarket strategies. Hence:
regulatory agency is negatively related to the
performance of a firm's nonmarket strategy. Hypothesis 4. A firm's experience in dealing
with government policy makers is positively
Nonmarket related to the performance of its nonmarket

Capabilities
strategy.
As argued earlier, the political market framework
provides one explanation forwhy nonmarket capa Another important dimension of transaction
bilities are particularly important in explaining costs in political markets is low transaction fre
firm nonmarket performance. Political markets typ
quency (Kaufman, Englander, & Marcus, 1993). Pol
ically suffer from higher transaction costs than do issues a firm come
icy affecting particular typically
economic markets, a difference that, we argue, cre
onto the political agenda only rarely (Kingdon,
ates a critical advantage for firms that have devel
1984). This temporal element implies that political
oped capabilities to mitigate these costs. Transac
markets are often discontinuous: interactions
tion costs in political markets arise, in large part,
from the potential opportunism of demanders and among demanders and suppliers are intense for a
time and then disappear for a longer period. In such
Parties may strike an agreement, yet the
suppliers.
a context, mitigating transaction costs through in
impossibility of or uncertainty surrounding judi
cial enforcement makes it difficult for the parties to tensive and repeated interaction is often not an
available option.
credibly commit to deals (Dixit, 1996; North, 1990;
Russo, 1992). However, we argue here that this characteristic
The existence of transaction costs does not mean, renders important another way by which firms can
however, that all firms are affected similarly. We develop transaction cost-mitigating capabilities: by
interact with gov learning from other firms' experience in similar
argue that firms that repeatedly
ernment policy makers gain an advantage in sus nonmarket settings. Studies on technological inno
taining trade in political markets in two ways. First, vation and geographic expansion strategies, for ex
existing research shows that the development of ample, have shown that firms learn from other
mutual trust, reputation, and cooperation is central firms in the same industry (Baum, Li, & Usher,
to solving commitment problems (Dyer, 1997; 2000; Jacobson, 1992; Macher & Henisz, 2004). A
Fukuyama, 1996; Hill, 1990; Jones, 1995). Such similar mechanism may enable firms to develop
attributes come from repeated interactions among transaction in political
cost-mitigating capabilities
demanders and (Williamson, 1994).
suppliers markets. Some of the heterogeneity among firms'
Firms that frequently engage with the government nonmarket performance is therefore likely to stem
thus have a chance to build trustworthy reputa from whether able to learn from
they have been
tions. Second, an important by-product of repeated others' Thus:
experiences.
interactions is the opportunity for firms to learn
from experience and to develop specific capabili
Hypothesis 5. A firm's opportunity to learn
from other firms' interactions with government

3 is positively related to the per
See for instance "Mario Monti's Parallel Universe," policy makers
Financial Times, November 6, 2002. formance of its nonmarket strategy.

Bergh
EMPIRICAL INVESTIGATION firms' decisions to file with the U.S. International

Trade Commission to obtain trade protection.
IndustrySetting Another advantage of using electric utility rate
Before discussing our empirical approach, we reviews for our empirical setting was that they
briefly outline the regulatory and political environ provide a good measure of the performance of a
ment of our selected industry and some of the reasons firm's nonmarket strategy (our dependent vari
why it provided a good setting for examining non able). As noted earlier, the lack of sufficiently
market strategy. To test our hypotheses, we focused detailed data has hindered management research
on the case of nonmarket strategy in the U.S. electric ers in empirically studying the performance as
utility sector. State agencies regulate the profit levels pect of nonmarket strategies. As part of their final
of utilities under a financial rate-of-return regime; rate review rulings, PUCs determine the financial
utilitiescan improve their financial performance by rate of return on equity (hereafter, the ROR) that
nonmarket strate a utility may earn and that determines allowed
achieving?through appropriate
rates of return. State regulatory agencies rate levels. Since, all else being equal, higher
gies?higher
RORs lead to higher profits, utilities prefer higher
(public utility commissions, hereafter, "PUCs") deter
mine the rate of return a utility is allowed to earn, and RORs. Although PUCs have a statutory duty to set
hence the final rates charged to consumers, through rates that are "just and reasonable," in practice
an administrative process, commonly termed a "rate they have considerable discretion to set rates and
review." A utility can file for a rate review whenever RORs within some implicit range.5 Utilities that
itwishes. Upon initiation of a rate review, a series of design effective nonmarket strategies may thus
com achieve higher RORs than other utilities. We
public hearings is held inwhich the utility and
interest and infor therefore used the rate of return as the basis for
peting groups present arguments
our measure of a utility's nonmarket
mation supporting their positions about justifiable perfor
rates of return and rate levels. At the end of this mance. This measure is also firm-specific: each
PUC commissioners make a final decision on ROR applies to a single utility. This fact allowed
process,
the rate of return for the utility and the rates that final us to overcome another common empirical prob
consumers lem for research on nonmarket strategy: since
pay.
often to all firms in an indus
The rate review process is characterized by an regulations apply
intense between try, it can be difficult to empirically assess the
informational exchange policy
the utility involved, and other interest effectiveness of a firm's individual strategy.
makers,
Third, the rate review process affords the opportu
groups (Hyman, 2000). Since providing informa
tion regarding policy consequences and alterna nity forboth the demand and supply sides of a polit
tives is a central characteristic of nonmarket strat ical market to have an influence on final policy out
comes. On the demand side, organized interest
egy (Hillman & Hitt, 1999), a utility's initiation of a
groups that are opposed to a utility's requests?large
rate review is a clear indication of the implemen
or industrial consumers, residential consumer advo
tation of such a strategy. At the same time, a utility
cates, or environmentalists, for a
is likely to engage in other nonmarket activities that example?have
its regulatory filing with the agency, right to participate in review hearings, to scrutinize
complement
utility expenditures, and to argue against rate in
such as gaining the support of the state governor
creases. Since PUCs must base their decisions on
and legislature (through lobbying, grassroots mobi
presented evidence, credible arguments from these
lization, coalition building, and financial campaign
contributions).4 groups can affect allowed RORs. On the supply side,
This industry context afforded a number of advan multiple regulatory and political institutions have a
potential role in rate reviews. Final decisions are un
tages for our empirical investigation. First, we were
der the jurisdiction of state PUCs. However, these
able to identify when firms engaged in a concerted
commissions are overseen by state legislatures, which
nonmarket strategy by observing when utilities filed
determine their budgets, can conduct hearings on
formal regulatory requests for rate reviews. By using
specific decisions, and can ultimately overturn PUCs
regulatory filings, we adopted the approach of other
nonmarket and Rehbein through new legislation. Further, state governors typ
strategy studies. Lenway
and Sch?ler for instance, considered ically appoint PUC commissioners, appointments of
(1991) (1996),
5
Allowed RORs have historically differed signifi
4
Data on state-level electoral contributions over utilities, states, and time. For instance, the
campaign cantly
from www.followthemoney.org demonstrate that electric highest ROR a state PUC allowed during 1980 was 16.80
utilities are important donors in political campaigns. percent, and the lowest was 12.50 percent.

fering another lever with which state politicians can estimated the following sample selection model,
exert pressure on PUC decisions. The attractiveness which incorporates a utility's decision to initiate a
of the political market for a utility is thus likely to be rate review (Greene, 2003; Heckman, 1979).
shaped by elected state politicians as well as by the rate review initiation decision:
Utility
regulatory agency.
Air = X1?1 + e1. (1)
Sample Initiation = 1 if Att > = 0 otherwise.

0,
After obtaining information of on the outcomes
all rate reviews initiated by the population of 190 (2)
investor-owned electric utilities during the period
PUC return on equity decision:
1980-92,6 we had a potential sample of 2,470 util
ity-year observations. After eliminating observa = =
(AROR\initiation 1) X2?2 + s2. (3)
tions for which data were missing, we were left
Correlation =
with 1,720 utility-year observations.7 The sample (s1,s2) P
included 491 rate reviews initiated by utilities.
In Equation 1, A 7rrepresents the expected change
in utility profits that would occur if a rate review
Methodology were implemented. Since a utility's decision rule,
as specified in Equation 2, is to initiate rate reviews
To test our hypotheses, we used a regression
only when Att is greater than 0, A7r is a latent
model of ROR decisions. However, since rate re
variable. X1 is a vector of variables including polit
views are not generated randomly, there is a poten
tial sample selection problem in using observed ical, institutional, and socioeconomic factors that
affect the attractiveness of the political market and
rate review information. Specifically, utilities will
thus capture a utility's expectation that a public
not initiate rate reviews if they expect PUCs to rule
If a utility initiated a rate review in a utility commission will increase its rate of return.
unfavorably. 3 estimates the change in the PUCs
Equation al
given period, then we did not observe the underly
lowed rate of return since the utility's last rate
ing regulatory environment. Normal ordinary least on observing a rate re
review, Af?Ofl, conditional
squares (OLS) regression techniques using only ob
view. X2 is also a vector of variables that includes
served rate review data would thus yield biased
measures of the political and regulatory environ
estimates of the impact of our explanatory variables
ment and other factors thatmight affect the change
on ROR decisions. To produce unbiased estimates,
in the allowed rate of return.
in the second part of our analysis we therefore the error terms of Equations
When 1 and 3 are
correlated?that is, p is not 0?simple OLS estima
6
These utilities comprise those operating in all U.S. tion of Equation 3 results in biased coefficients. To
states except Alaska and Nebraska. We concentrated on correct for selection bias, we thus used the Heck
1980-92, since utilities initiated rate reviews then in man full-information maximum-likelihood estima
to rising costs. After 1992, as costs to tion procedure from STATA. This procedure
response began yields
decline, PUCs began to initiate rate reviews with the aim unbiased estimates of ?2 coefficients.
of reducing rates. Since our was to ex
utility objective
amine utility strategy, we thus focused on the prior
period. Data and Measures
7
Specifically, to measure our dependent variable
in allowed ROR), we needed a baseline measure Dependent variable. To measure nonmarket
(change
of allowed ROR. Thus, we eliminated observations from performance, we calculated the change in allowed
the period before each utility's firstrate review. We also rate of return (AROR) that had occurred since a
eliminated observations ifwe were missing information rate
utility's previous review. We used change in
on the allowed rate of return for a firm, since the absence
ROR rather than its absolute level since doing so
of that information made it impossible to calculate the allowed us to control for constant firm-level factors
change in allowed ROR. The need for a baseline and the thatmight influence the absolute rate. We obtained
missing data on allowed ROR resulted in the exclusion of the rate review data from a private firm, Regulatory
311 observations. We also eliminated 384 observations
an
Research Associates, that tracks PUC decisions
data for measuring independent variable, util
lacking
to assess we cross-checked a
ity revenue/PUC budget, and 22 observations
and, accuracy, sample
lacking
data on another variable, market share. Fi of rate review results with data available in annual
independent
volumes of the National Association of Regulatory
nally, a lack of data on three utilities resulted in 33
additional observations being eliminated. Utility Commissions (NARUC).

Independent variables. Hypothesis 1 concerns legislature rivalry were coded 1 if rivalry was in
interest group rivalry. We used three variables to tense and 0 otherwise. We used dummy rather than
capture different sources of potential demand-side continuous variables since the underlying distribu
competition from organized interest groups. Con tions of governor vote and legislature party major
sumer advocate was a measure of the degree to ities were not normal but highly skewed. We col
which residential utility consumers in a state were lected this information from annual volumes of The
organized. Thirty states have created consumer ad Rook of the States, published by the Council of
vocacy offices charged with the express purpose of State Governments.
representing residential utility consumer interests Hypothesis 3 concerns regulatory agency re
before state regulatory agencies and courts (Hol source bases; PUCs with greater resources are less
burn & Vanden Bergh, 2006). Consumer advocates, dependent on information provided by utilities.
which have public funding and statutory power to Again, we used several measures. Our first, PUC
participate in rate review procedures, can provide budget per capita in a state, was a measure of fi
strong opposition to utility requests for rate nancial resources. Second, we constructed a mea
increases (Holburn & Spiller, 2002). The variable sure of PUC commissioner experience since expe
consumer advocate equaled 1 if a consumer advo rience may partially substitute for financial
cacy office existed in a given state in a particular resources: Average tenure of commissioners was
year and 0 otherwise. Rivalry can also come from equal to the sum of each commissioner's tenure in
industrial consumers who, because they have years divided by the total number of commission
higher average levels of consumption than residen ers on a focal PUC. We expected that more experi
tial consumers, have stronger incentives to orga enced commissioners would have better informa
nize. Industrial consumers, a time-varying variable, tion andinsights regarding utility rate review
was to the industrial percentage share of elec
equal requests. We obtained annual information on PUC
tricity consumption in a given state. Data on elec budgets and the identities of PUC commissioners
tricity consumption by consumer sector were from annual reports of the National Association of
obtained from the Energy Information Administra Regulatory Utility Commissioners, annual volumes
tion. Finally, we used Sierra Club membership to of The Rook of the States, and the Web sites of
capture the extent to which a state population par individual PUCs. Third, we allowed for PUC re
ticipated in environmental and other activist non sources to vary relative to individual utilities as
governmental organizations. The Sierra Club is the well as in an absolute sense; a PUC with a small
largest environmental NGO in the United States. budget is less dependent on the utility if the utility
Such groups havehistorically been
particularly itself has minimal resources. Hence, utility reve
active against utilities regarding the siting of new nue/PUC budget was the dollar value of utility
power generation plants and the environmental electricity revenues within a state divided by the
impacts of existing facilities. To normalize mem PUC budget in each year. Information on utility
bership levels across the states, we calculated Si revenues was gathered from federal utility filings
erra Club membership as the total number ofmem available through the Energy Information
bers divided by a state's population (in thousands). Administration.
Annual information on state membership was pro To capture a utility's experience in dealing with
vided directly to us by the Sierra Club. policy makers, which Hypothesis 4 concerns, we
Hypothesis 2 concerns political rivalry. We relied on two related measures. In the selection
based two dummy variables on the margin of win equation, we created the variable cumulative rate
ning votes in the most recent state gubernatorial reviews by utility, which is equal to the total num
and legislative elections; these were proxies for the ber of rate reviews a utility has experienced at a
degree of rivalry among elected politicians. For the given time. In the regression equation, we created
executive branch (governors), we considered ri recent rate review, a dummy variable set equal to 1
valry intense
if the margin of electoral victory be if a utility had experienced a rate review in the
tween the winning and second-place candidates previous three years and 0 otherwise. We differen
was less than 5 percent. In this case there is likely tiated between the initiation of a rate review and
to be intense political competition during the next performance in the review since we anticipated
electoral cycle. For the legislative branch, given the that total experience in a variety of regulatory set
importance of party control of legislatures, we con tings would affect a utility's decision to initiate. Its
sidered rivalry intense if the margin of control by performance in the rate review, however, would be
the majority party (measured by the number of more closely related to its recent experience, since
seats in the combined upper and lower chambers) the characteristics of the regulatory environment
was less than 5 percent. Thus, governor rivalry and
change over time.

To address Hypothesis 5, concerning other firms' in their RORdecisions. Elected PUC was a dummy
experiences with policy makers, we used other variable equal to 1 in states where PUC commis
a dummy variable set equal sioners were elected and 0 otherwise. PUC commis
firms initiating reviews,
to 1 if other utilities in the state initiated rate re sioners were elected by the voting population in
views with their PUC in the previous year and 0 ten states and appointed by the governor in other
otherwise. The variable captured potential utility states. Prior research suggests that elected PUCs
learning from observing other utilities' experience place greater weight on consumer welfare (Besley &
with the commission. Coate, 2003). Details on commissioner selection
were obtained from The Rook of the States. Simi
larly, the variable Republican governor and legisla
Control Variables
ture equaled 1 if there was unified Republican con
We controlled for a number of factors that might trol of the branches of state government and 0
affect a utility's performance in the rate review otherwise. Use of this variable
captured the poten
process as well as its decision to initiate a rate tial impact of ideological factors (as proxied by
review. Interest rates on treasury securities enter political party) on regulatory policy and utility
into a PUCs decision on the allowed ROR since strategy.
these are a benchmark for measuring the cost of Finally, we also measured the market share for a
in interest rate, measured in per utility as the total megawatt hours (MWh) of elec
capital. Change
centage points, was the difference between the in tricityit provided divided by the totalMWh pro
terest rate on ten-year Treasury bills at a given time vided by all utilities in its state. If a utility is a
minus the interest rate at the time of the last rate major player within a public utility commission's
review. Change in average fuel cost, the percentage jurisdiction, then that utility's information is likely
change in a utility's average fuel costs (on per Btu
a to be more valuable to the PUC than information
basis) since its last rate review, was driven mainly from smaller utilities. Market share thus measured
by external market forces. Increases in the cost of the influence of a utility relative to other utilities.
utilities' fuel purchases, such as those that oc Tables 1 and 2 give descriptive statistics and
curred during the early 1980s, directly reduce util correlations for the variables. 1 provides
Table sta
ity profits, thereby increasing the probability that tistics for variables in the full sample of
included
utilities will initiate rate reviews.8 In the selection utility-year observations used in the rate review
equation, we also controlled for the absolute level initiation (selection) model, and Table 2 provides
of fuel costs; since absolute costs are inversely re statistics for variables included in the change in the
lated to profits, we expected a positive relationship allowed rate of return (regression) model.
between absolute costs and the probability that
utilities initiated. We measured average fuel cost as RESULTS
the average price of fuel per Btu purchased by
electric utilities within a state. Fuel cost data were We begin by discussing the results of the selec
from the Energy Information Administration. To tion-corrected change in ROR regression model. Ta
control for varying economic conditions across the ble 3 shows the results of our model estimated with
states, we included a measure of the change in per state fixed effects. The statistically significant Mills
capita income (lagged one year), which was equal ratio coefficient supports our empirical approach:
to the annual percentage change in per capita in we can reject the null hypothesis at the 1 percent
come in a state; voter pressure on utility rates may level of confidence that there is no sample selection
be inversely correlated with recent economic problem. With only one choice for utilities (initiate
trends. We these data from the a rate review or not), the positive coefficient on the
growth gathered
Bureau of Economic Analysis. Mills ratio implies that a positive correlation exists
We included additional political and institu between the decision to initiate?and therefore to
tional variables that might influence the weight engage in a nonmarket strategy to change an exist
that PUCs put on utility versus consumer interests ing regulation?and the performance of a utility in
the rate proceedings (Dolton & Makepeace, 1987).
In other words, we found good evidence that utili
8 ties used the rate review initiation process strategi
Some states adopted automatic fuel adjustment
cally. Among control variables, it is also worth
clauses (FACs) during the 1980s, which allowed utilities
to pass through fuel costs without formal rate review. noting that the variables change in interest rate and
in income were significant and
However, since such clauses rarely allowed utilities to change per capita
pass through 100 percent of the cost increases, fuel-cost positive. As expected, though not directly related
triggered
rate reviews were not completely eliminated. to the political markets logic, changes in the cost of

.04
17
16 .03
.10
15 .15 .11
-.11
.10 .05 .19 .03

14
.04 .04
13 .07 .02 .19
-.05 .11 -.06.11

12 .09
-.36
.01 .12 .04 .05 .01 .31

11 .07
.14 .06 .06 .15 .21 .04

10 .22 .17
.07 .60 .06 .00 .05 .01 .12

.12 .05
Variables
andSummary
StatisticsRate
Review InitiationEstimation3
TABLE
1 for
Elected
PUC
0.35
0.14
-.10
8.
-.01
-.05
-.28
-.07
-.27
.08 Market
0.26
0.29
share
-.03
-.06
11.
.00
.03
.06
.13
-.05
.25
15.Average
fuelcost1.68-.07
0.81
-.03 .14.16-.15 -.27 -.07
.09
budget
PUC
2.00
1.47
7.
.07
.05
.12
-.14
-.05
-.10
10. Cumulative rate reviews .16 Change

3.82
1.96
interest
in
1.89
12.
-1.33
.09
.00
.02
rate
.04
-.02
-.10
.07
.00 -.17
Change
income
in
0.03
0.06
17.
capita
.00
-.01
.12
-.13
-.09
per
.01
.00
.00 .18 .00 .07 .11 -.16
14.Change
inaveragefuelcost -.1118.06
.06
-4.44 .12-.01-.02.01 .15 .05
budget
Utility
130.91
105.38
9.
revenue/PUC
-.05
.14
.00
.05
-.31
-.06
.09
18. Mean
s.d.Other
5.6. Legislature
Governor rivalry
firms
rivalry
.02 -.04
0.410.30
13. .10.15
initiating
Average
tenureof-.04.16 -.08
commissioners
reviews
2.59 -.15-.06-.17
0.46
-.03
.14
.00 .21 .45 .09 -.02 .01 .08 .03 -
0.10 .06-.05.12 3.63
0.22 0.69
16. Republican governor and 0.10 0.30 -.02 -.11 .06 -.16 .11
legislature
1 n= 1,720.
"PUC" is
utility commission."
"public
4. Sierra
Club1.50
1.07 -.02-.12-.20
membership
3. Industrial 0.08
0.27
consumers-.02-.11
2. Variable 0.59
Consumer 0.49 -.01
advocate
1. Rate
0.29 0.45
review

16 .05
.15
.02
15
14
-.09.35
.01 .01 .04
-.15
13
-.12
12
.08
-.13 .02
-.11
.00
.43
.05 .30
.02
.12
.31
11
10
Variables
and
Summary TABLE
2 in
Statistics,
Allowed Rate
ofReturn
Estimation8
Change
Market
0.25
0.31
share
11.
-.15
.05
.10
.00
.16
-.07
.01
-.01
.25
.59
Elected
PUC
0.28
0.09
8.
-.01
-.21
-.04
-.15
-.01
.15
7. PUC
budget
1.78-.181.23
-.11 Recent
10.
0.86
rate
0.35
-.06
-.19
review
-.08
.02
.46-.03Change
.06
.01
.16 interest
in
1.81
-.12
12.
-1.20
-.04
.04
.03
.01
.55
rate
.02
-.06
.02
.36 .01
-.19
Change
income
in
0.03
0.07
16.
capita
.43
-.17
.01
-.08
.13
per
.04
.08
.21
17.Other
firms 2.60
Average
3.32
of
tenure
commissioners
13.
-.12
-.16
.04
-.13
.08
-.06
-.04
-.09
.36
.29Change
in
14.
fuel
16.60
-2.63
cost
-.18
-.03
.43
-.11
.01
.03
.20
.08
average
.18
initiating.13
reviews 0.79.16
.01 .00.00.14 .02 -.20 -.15 -.13 .27
9. Utility revenue/PUC 0.40 -.06
s.d.budget 116.26 .04
-.03 .20 -.05 -.02 -.33
in
is
"PUC"
allowed
"change
of
"AROR"
"public
rate
utility
491.
return."
commission.'
1n
=
135.45
legislature
Republican
0.29
0.09
and
15.
-.14
-.10
.16
.04
-.01
-.03
governor
.27
.03
.03
6. LegislatureGovernor
0.42
0.23
5.
rivalry
-.06
.05
-.02
.08
rivalry
.070.27 .16 .16 .03 .22
1. Mean AROR
1.34 0.08
-0.47
Club
Sierra
1.00
1.46
membership
4.
-.15
-.09
-.22
3. Industrial
Consumer0.580.27
2.Variable 0.49 -.08 0.08 .09 -.13
advocate
consumers

Bergh
TABLE 3 the selection effect, we could not interpret the co

Results of Heckman Selection-Corrected efficients as straight marginal effects. We have thus
Regression Analysis for Change in Allowed included Table 4 to present selection-corrected
Return on Equity since Last Rate Reviewa marginal effects for each of the statistically signifi
cant variables. A marginal decrease in demand-side
and Variables b (s.e.)
Hypothesis
rivalry, as measured by Sierra Club membership,
was expected to increase the ROR by six basis
1: Interest group
Hypothesis
rivalry points. The degree of rivalry generated by activists,
Sierra Club membership -0.38* (0.18) then, appears to be an important factor in the abil
0.29 (0.31) ity of utilities to achieve favorable PUC decisions.
Consumer advocate
Industrial consumers -0.02 (2.68)
This result is in line with previous research, which
Hypothesis 2: Political rivalry suggests that activists constitute a particularly dif

Legislature rivalry 0.57* (0.28) ficult threat for firms to handle (Bonardi & Keim,
Governor rivalry -0.09 (0.13) 2005). Our other demand-side rivalry variables,
however, did not display significant coefficients,
Hypothesis 3: Regulatory agency
perhaps owing to measurement challenges. The
resources
PUC budget -0.20+ (0.12)
consumer advocate dummy variable, for instance,
Average tenure of -0.08* (0.04) might have been too coarse to capture the strength
commissioners of consumer opposition. Finer-grained data, such
Utility revenue/PUCbudget 0.00 (0.00) as the budgets of consumer advocate organizations,
4: Firm's
were unavailable. A potential explanation for the
Hypothesis experience
Recent rate review 0.82** (0.16) lack of significance of industrial consumers is that
powerful industrial consumers, although present,
5: Other firms'
Hypothesis did not in fact compete against utilities on ROR
experience
-0.03 decisions?perhaps in return for utility support on
Other firms initiating reviews (0.15)
other policy dimensions, such as the rate structure,
Control variables where industrial consumers compete against resi
in per capita income 8.91**
Change (2.01) dential consumers.
Change in interest rate 0.23** (0.04)
The results suggest also that rivalry among poli
Change in average fuel cost 0.02** (0.00)
Market share 0.06 ticians shapes PUC decisions, which provides sup
(0.26)
Republican governor and 0.30 (0.25) port forHypothesis 2. The coefficient on legislature
legislature rivalry was positive and statistically significant at
Elected PUC -0.07 (0.95) the 5 percent level when we included state fixed
Constant -1.37 (1.07)
effects in the model. As for the effects ofmargin of
Mills ratio 0.36** (0.13)
State dummies Yes
n 491
Log pseudo-likelihood -1,577.92 TABLE 4
a Marginal Effects from Heckman
"PUC" is "public utility commission."
+ Selection-Corrected Regression Results for
p < .10
* <
.05 Change in Allowed Return on Equity
p
**p
< .01 since Last Rate Review
Marginal
Statistically Significant Variable Effects3
financing an impact on the change in
should have
ROR. Similarly, annual fluctuations in state eco Sierra Club membership -0.06
nomic conditions were positively correlated with Legislature rivalry 0.18
increases in utilities' allowed rates of return. PUC budget -0.02
tenure of commissioners -0.01
Turning to our key variables, we found good Average
Recent rate review 0.22
statistical support overall for our hypotheses. First, in per capita income 0.37
Change
regarding demand-side rivalry (Hypothesis 1), the Change in interest rate 0.08
coefficient on Sierra Club membership is negative Change in average fuel cost 0.002
and statistically significant at the 5 percent level, a

effects were calculated at the mean of each con
suggesting that lower levels of interest group ri
Marginal
tinuous independent variable and for discrete changes in each
valry (competition from organized interest groups)
dummy variable. We present the marginal effects for the model
led to positive changes in the ROR for the utility. that includes state dummy variables. "PUC" is "public utility
Note, however, that, because of the nonlinearity of commission."

control by the majority party, we found that when (Hypothesis4), we found a positive and statisti
rivalry in the legislature in a utility's state was cally significant coefficient on recent rate review,
intense, the utility could expect to benefit from an which provided strong evidence that utilities with
18-basis-point increase in ROR. Interestingly, when recent rate review experience tended to perform
we computed some robustness checks by using al better in subsequent rate reviews. At the margin,
ternative constructions of the dummy variable with increasing the value of recent rate review increased
different cut-off points, we found that the impact of a utility's ROR by 22 basis points. On the other
legislative rivalry was even larger (and significant) hand, we did not find support for Hypothesis 5,
when narrowing down the cut-off point. With a 2 which argues that firms can learn by observing
percent cut-off, for instance, the estimated coeffi others' past nonmarket interactions.
cient in the regression doubled.9 On the other hand, Among the remaining control variables, the vari
with a cut-off of 20 percent, the coefficient was ables representing a Republican governor and leg
correctly signed but not significant. This check islature, an elected PUC, and market share were
supported the idea that very strong levels of polit signed as expected but not significant. We experi
ical rivalry generate especially positive situations mented with other control variables that might
for firms' nonmarket strategies. have affected the ROR, such as utility operating
Rivalry among elected political candidates thus efficiency and the concentration of utilities within
appears to create an opportunity for utilities, a par a state, but did not find evidence of an impact.
ticularly well organized interest group, to "pur Our

empirical specification also generated in
chase" regulatory policies (i.e., higher allowed sights into the reasons why firms proactively re
RORs) through additional campaign contributions, quest a change in regulatory policy. Table 5 pre
grassroots mobilization, or other politically valu sents the selection results, in which the dependent
able resource transfers. Legislatures can induce
PUCs to cooperate on rate review decisions by TABLE 5
threatening budget cuts or by supporting legislative
Results of Heckman Selection-Corrected
proposals authority. We did not
that constrain PUC
for
Regression Analysis
find any evidence, however, that rivalry among
Utility Initiation of Rate Reviewa
gubernatorial candidates (our variable governor ri
valry) influenced PUC decisions. This absence of Variables b (s.e.)
evidence may reflect the weaker ability of the ap

is largely the preserve
Sierra Club membership -0.13** (0.04)
pointments process, which Consumer advocate -0.24** (0.09)
of governors, to immediately impact PUC commis Industrial consumers -0.29 (0.46)
sioner decisions.
Going beyond rivalry factors, we observed that Legislature rivalry -0.19 (0.13)
Governor 0.10
regulatory agency resource affected rivalry (0.09)
dependence
the ability of firms to achieve preferred policy rul PUC budget -0.09** (0.03)
ings. Measures of both financial and experiential Average tenure commissioners -0.03* (0.02)
PUC resources were important (Hypothesis 3). The Utility revenue/PUC budget -0.00 (0.00)
negative and statistically significant coefficient on Cumulative rate reviews 0.19**

by utility (0.02)
PUC budget indicated that utilities' RORs were neg
atively affected as regulators' financial resources Other firms initiating rate reviews 0.44** (0.09)
increased. Similarly, the negative and statistically
in per capita income 6.49** (1.24)
significant coefficient on the average tenure of com
Change
Change in interest rate -0.06** (0.02)
missioners supported the contention that greater
Change in average fuel cost 0.01** (0.00)
PUC experience is detrimental to utility perfor Market share 0.60** (0.23)
mance. The economic significance of these effects, Republican governor and legislature -0.02 (0.13)
however, appeared to be less important than those Average fuel cost 0.14** (0.06)
Elected PUC -0.28* (0.15)
of rivalry. Marginally increasing PUC budget and Constant -1.70** (0.30)
the average tenure of commissioners reduced the
ROR by 2 and 1 basis points, respectively. State dummies Yes
Turning to firm experience with policy makers Wald test of independent equations (^J 9.05
n 1,720
Reviews correctly classified by model 73%
9 1
A test on a continuous measure of "PUC" is "public commission.'
Shapiro-Francia utility
*
the legislative rivalry variable also demonstrated that it p < .05
was not normally distributed. **p
< .01

variable was whether or not a utility initiated a rate TABLE 6

review. Table 6 presentsthemarginal effect of each Marginal Effects from Heckman Selection
statistically significant variable on the probability Corrected Regression for
Results
of initiation. The full model presented in Table 5 Utility Initiation for Rate Review
correctly classified 73 percent of the cases, suggest
in capturing the Marginal
ing this model performed well
Statistically SignificantVariable Effects3
initiation dimension of the utilities' nonmarket
strategy. Consumer advocate -0.08
In general, the results on individual initiation Sierra Club membership -0.05
model variables with PUC budget -0.03

displayed strong consistency
tenure of commissioners -0.01
the pattern of results in the change in ROR model. Average
Cumulative rate reviews by utility 0.06
Increased demand-side rivalry with other interest Other firms initiating rate reviews 0.13
groups appeared to dampen the incentives of util Elected PUC -0.09
ities to initiate rate proceedings with their state Market share 0.20
PUCs. As itwas in the change in ROR model, Sierra Change in interest rate -0.02
in average fuel cost 0.002
Club membership was significant and negative in Change
Average fuel cost 0.05
the individual initiation model, but this time con
sumer advocate displayed a similar result. At the a
Marginal effects were calculated at the mean of each con
margin, Sierra Club membership and the presence tinuous independent variable and for discrete changes in each
of a consumer advocate reduced the probability dummy variable. We present the marginal effects for the model
that includes state dummy variables. "PUC" is "public
that a utility initiated by nearly 5 and 8 percent, utility
commission."
respectively. Rivalry created by other demanders of
public policy therefore seemed to be an important
factor in a utility's analysis of its nonmarket envi increased by nearly 20 percent. This result is con
ronment and in its decision to implement a non sistent with previous studies showing that firm size
market strategy. is a determinant of the decision to engage in a
Likewise, the negative and statistically signifi nonmarket strategy (Masters & Keim, 1985;
cant coefficients on PUC budget and the average Munger, 1988; Sch?ler, 1996; Zardkoohi, 1985).
tenure of commissioners both suggested that as a Finally, the control variables were generally sig
regulator's dependence on a firm for informational nificant and signed as expected.
resources declines, the attractiveness of the politi
cal market for the utility also falls. At themargin, as
DISCUSSION
PUC budget and the average tenure of commission
ers increased, the probability of a utility initiating a In this study, we set out to develop and test a
rate review declined by about 3 and 1 percent, model of what determines the performance of a
respectively. These results were consistent with firm's nonmarket strategy in the context of a spe
Hypotheses 1 and 3. We did not obtain statistical cific regulatory or political issue. Building on the
significance, however, on the political rivalry vari political markets framework, according to which
ables (Hypothesis 2). public policies arise from the interaction of de
Firm-level variables also performed well in the manders and suppliers of such policies, we argue
rate review initiation model. As utilities accumu that nonmarket performance is influenced both by
lated knowledge and experience about the rate re the characteristics of a firm's regulatory-political
view process, as measured
by cumulative rate re environment and by the internal capabilities the
views by a utility, the probability of initiating a firm has developed over time. More precisely, we
review increased by 6 percent. Additionally, there hypothesized?and supported empirically, in the
was evidence of a spillover effect from other utili context of U.S. electric utilities?that the rivalry
ties: the variable for other firms initiating rate re created by competing demanders of public policies
views, statistically significant at the 1 percent level, (such as environmental activists), as well as the
was estimated to increase the likelihood of a utility resources of the regulatory agency involved, had a
triggering a rate review by nearly 13 percent. This negative impact on the firm's ability to obtain reg
result is similar to that of Hersch and McDougall ulatory approval for higher profit levels. On the
(2000), who found that in the U.S. automobile in other hand, we found that the rivalry among
dustry the major firms' levels of political activity elected politicians supervising policy implementa
were related to the political activities of their rivals. tion had a positive impact on regulatory rulings
Similarly, as market share increased, the proba favorable toward the firm. Last, we found that the
bility that a utility had initiated a regulatory review firm's previous experience with regulators through

1224 Academy of Management Journal December
making prior regulatory filings played an important mation to the regulatory agency and participating
role in explaining the performance of its nonmarket in public hearings, but it does not allow us to
strategy. identify the extent of a utility's investment in these
We make several contributions to the existing activities. Although we recognize that this is a
literature on nonmarket strategies. First, we pro shortcoming of our study, the positive findings in
vide a general model of firms' nonmarket perfor the empirical model of the utility's initiation deci
mance that integrates different aspects examined in sion suggest the dichotomous measure is nonethe
previous studies, including the attractiveness of less a reasonable strategy indicator.
political markets and firms' nonmarket capabili A related limitation is that we were unable to
ties. The literature so far has remained scattered, examine inmore detail the design of utilities' non
offering little focus on nonmarket performance and market strategies outside regulatory settings (Han
disparate theoretical perspectives. Lord (2000), for sen & Mitchell, 2000; Schuler et al, 2002); detailed
instance, presented the results of an interesting sur data on utility campaign contributions, lobbying,
vey of U.S. companies on the impact of various and mobilization of grassroots support for legisla
nonmarket activities (electoral campaign contribu tors were unfortunately not available for our sam
tions, informational lobbying, advocacy advertis ple. We do know, however, from aggregate state
ing, and constituency building) but did not provide level campaign contribution data (available from
any insights about the factors that affected their www.followthemoney.org) that electric utilities are
performance. We
believe that the concept of polit significant contributors to political candidates for
ical markets has the potential to provide such an state government offices. This suggests that future
integrative framework. As argued in this article, research incorporating multiple dimensions of firm
elements from economics, from the resource-based nonmarket strategy is warranted.
view of the firm, and from resource dependence Third, within the political markets framework
theory can be integrated into the framework to pro we provide a better understanding of the impact on
vide a comprehensive view of nonmarket perfor a firm's nonmarket performance of regulatory agen
mance, as well as a basis for future research. The cies responsible for policy implementation. Firms
framework, by delineating the conditions under a
in wide variety of industries, including agricul
which nonmarket strategies are likely to be effec ture, pharmaceuticals, and utilities, are subject to
tive, also provides guidelines formanagers who are industry-specific agency rule making. Many other
assessing whether to implement such strategies. firms are subject to functional regulations that cross
The second major contribution of our work is to industry borders, concerning issues such as work
provide unique empirical evidence supporting the place safety, labor standards, and environmental
theoretical validity of the political markets frame impact. An important step in our hypothesis devel
work and its implications for nonmarket strategy opment was depicting regulatory agency decision
performance. The data requirements for assessing making in general in the context of the broader
performance are challenging. Researchers must ob institutional environment, which includes courts
tain data relating to (1) an identifiable, specific and legislatures that have the ability to reverse er
political or regulatory issue, (2) the implementation rant agency actions. Agencies operate under differ
of firms' nonmarket strategies, and (3) a measure of ent incentives and constraints than elected politi
the policy outcome. Most existing studies of non cians. We argue specifically that regulatory
market strategy have used data on aspects 1 and 2 agencies are constrained in their decisions by pro
only, which precludes investigation of perfor cedural requirements: agencies need resources to
mance issues (e.g., Lenway & Rehbein, 1991; obtain information and to justify their rulings in
Sch?ler, 1996). Our data on electric utility rate order to avoid judicial override. We find strong
reviews provide good information on both a spe evidence that better funded, more experienced
cific political-regulatory issue, regulated levels of agencies are better positioned to counter the policy
1, and changes that firms claimare required.
utility profitability, satisfying requirement
on a with the results on political rivalry, our
good information policy outcome, regulatory Together
decisions on rate of return, satisfying requirement findings suggest that the ability of firms to gain
3. Our data satisfying requirement 2, which came favorable policy rulings in regulatory arenas?as
a
from dichotomous measure (presence or absence compared to legislative arenas?is complex, requir
of a utility-filed formal request for policy review) is ing firms to operate in multiple institutional envi
coarser-grained than the ideal measure of the im ronments. To be successful in regulatory arenas,
of a nonmarket strategy would be. firms may additionally need to gain the support of
plementation
Our measure indicates utility engagement in vari the legislative and executive bodies that monitor
ous nonmarket activities, such as providing infor the decisions of agencies and may subsequently

"punish" the agencies for errant decisions. Alterna this is a promising avenue for future research, as
tively, firmsmay be able to use these political chan are broader questions related to the impact, and
nels strategically to indirectly pressure agencies to source, of firm capabilities in achieving favorable
implement favorable rulings (Holburn & Vanden policy outcomes.
Bergh, 2004).
Our fourth contribution is to provide new empir
Limitations and Avenues for Future Research
ical results consistent with the notion that firms
can develop internal capabilities that improve their Although we believe our results are encouraging,
nonmarket performance. Existing research shows a number of other theoretical and empirical limita
that the prior experience of firms' board members tions in our analysis call for further research on the
in political institutions is associated with better topic of nonmarket strategy performance. One po
overall firm performance
(Hillman et al., 1999).
In tential shortcoming is that our hypotheses, albeit
the nonmarket we were in context a
extending capabilities literature, generic, tested the of single in
argue that repeated interactions between firms and dustry, raising questions about the generalizability
policy makers are likely to be an important mech of our findings to other settings. We might expect to
anism for developing nonmarket capabilities, since find that the role of firm nonmarket capabilities, for
they provide an opportunity for firms to learn from instance, is less significant in industries that are
experience and to establish organizational routines. less heavily regulated than the utilities sector,
In addition?and critically?repeated interactions where firm-regulator interactions are relatively fre
enable firms to establish a reputation for credibility quent. Utilities also have a unique ability to initiate
with government actors, which is essential for over policy change through the rate review process;
coming the high transaction costs of exchange in without such rights, firms in other industries may
political markets. Although we found that firms find itmore difficult to gain access to policy makers
with greater experience in interacting with regula and to establish political markets, making political
tory agencies through prior rate reviews did indeed strategies less effective. A further characteristic of
achieve better policy outcomes, we did not find our research design is that we measured firm-level
evidence that firms achieved similar results by ob regulatory policy outcomes and firm-level strate
serving the experiences of other firms, implying gies aiming for private benefits of regulation. Al
that nonmarket capabilities are not easily imitated though we regard this measurement focus as a
(Keim & Baysinger, 1988). It is interesting to note, strength of our analysis in identifying performance
however, that utilities were more likely to initiate drivers, it does mean that we did not explore col
rate reviews when other utilities also initiated lective action problems within an industry. If reg
them. This finding is consistent with Oliver's ulations provide public rather than private benefits
(1991) proposition that regulated firms can obtain a to firms within an industry, the costs of organizing
degree of legitimacy by conforming to institutional collective action could, all else being equal, reduce
processes. nonmarket strategy performance.
Other types of capabilities have been extensively From a theoretical perspective, we did not dis
studied elsewhere (Ethiraj, Kale, Krishnan, & tinguish between different types of nonmarket
Singh, 2005; Henderson & Cockburn, 1994; Maka strategy in our hypotheses. Different types of strat
dok & Walker, 2000; Silverman, 1999), yet as far as egy are likely to perform differently in various en
we know, our study is one of the first empirical vironments (Hillman & Hitt, 1999). For instance,
attempts to directly measure nonmarket capabili theways inwhich firms attempt tomitigate interest
ties. Our study also raises a related question: if group competition will not be the same as the
firms learn from their own experiences in a partic methods by which they gain the support of elected
ular institutional environment, can they redeploy legislators. The incentives and objectives of these
these nonmarket capabilities to other institutional two groups vary, and firms will adapt their non
settings? We tentatively explored this issue using market tactics accordingly. There is thus scope for
our data by considering parent company experi future work to consider a finer-grained measure of
ence for those utilities that belonged to holding nonmarket strategy and to consider the demand
companies. Our initial results (unreported here) side and supply-side conditions inwhich each type
suggested that a holding company's rate review will be more or less effective. Another potential
experience in other states was correlated with im route for developing the political markets frame
proved rate review outcomes in a focal state, im work is to examine the interactions between and
plying that firms may learn some generic lessons within demand- and supply-side factors. Does in
about interacting with policy makers through their terest group competition, for example, have a more
experiences in different jurisdictions. We believe powerful effect when political rivalry is also

strong? How does the impact of regulatory exper former monopolies. Strategic Management Jour
tise depend on political rivalry? We believe that nal, 25: 101-120.
addressing these types of questions would provide Bonardi, J. P., Hillman, A., & Keim, G. 2005. The attrac
important additional theoretical insights into the tiveness of political markets: Implication for firm
attractiveness of political markets. strategy. Academy of Management Review, 30:
397-413.
Bonardi, J. P., & Keim, G. 2005. Corporate political strat
Conclusion egies forwidely salient issues. Academy ofMan

agement Review, 30: 555-576.
Despite these and other limitations, our study
new Buchanan, J., & Tullock, G. 1962. The calculus of
con
offers insights into the factors that affect the sent: Logical foundations of constitutional democ
ability of firms to succeed when they engage in
racy. Ann Arbor: University ofMichigan Press.
nonmarket strategies. In particular, we argue that
T. J., & Brown, R. L. 1995. Pollution as a
both the external environment, which we concep Dean, regulation
tualize as a political market involving demanders barrier to new firms entry: Initial evidence and im
plications for future research.
and suppliers of public policies, and the internal Academy of Manage
ment Journal, 38: 288-303.
characteristics of firms both matter significantly to
de Figueiredo, J.M., & Silverman, B. 2006. Academic
nonmarket performance. Using data from U.S. elec
tric utilities, we find good empirical earmarks and the returns to lobbying. Journal of
support for
Law and 49: 597-626.
this thesis, even though much work remains to Economics,
be done. de Figueiredo, J.M., & Tiller, E. H. 2001. The structure

and conduct of lobbying: An empirical analysis of
corporate lobbying at the Federal Communications
Commission. Journal of Economics and
REFERENCES Manage
ment Strategy, 10: 91-122.
Ansolabehere, S., de Figueiredo, J.M., & Snyder, J.M. Dixit, A. 1996. The making of economic policy: A trans
2003. Why is there so littlemoney in U.S. politics?
action-cost perspective. MA: MIT Press.
Cambridge,
Journal of Economic Perspectives, 17: 105-30.
Dolton, P. J., & Makepeace, G. H. 1987. Interpreting
sam
Aranson, P. 1990. Theories of economic From
regulation:
ple selection effects. Economics Letters, 24: 373
clarity to confusion. Journal of Law and Politics, 6: 379.
247-286.
resources com
Dyer, J.H. 1997. Effective interfirm collaboration: How
Barney, J. B. 1991. Firms and sustained
firms minimize transaction costs and maximize
petitive advantage. Journal ofManagement, 17: 99
transaction value. Strategic Management Journal,
120.
18: 553-556.
Baron, D. 2001. Theories of strategic nonmarket partici
Eggertson, T. 1990. Economic behavior and institutions.
pation: Majority-rule and executive institutions.
New York: Cambridge University Press.
Journal of Economics and Management Strategy,
10: 47-90. Ethiraj, S., Kale, P., Krisfman, M. S., & Singh, J.V. 2005.
Where do capabilities come from and how do they
Baron, D. 2003. Business and its environment. Upper
matter? A study in the software services industry.
Saddle River, NJ: Prentice-Hall.
Strategic Management Journal, 26: 25-45.
Baum, J.A. C, Li, S. X., & Usher, J.M. 2000. the
Making
next move: How experiential and vicarious learning Fukuyama, F. 1996. Trust: The social virtues and the
creation of property. New York: Free Press.
shape the locations of chains' acquisitions. Admin
istrative Science Quarterly, 45: 766-801. Getz, K. 1997. Research in corporate political action:
as an Integration and assessment. Rusiness and
Baysinger, B. 1984. Domain maintenance objective Society,
of business political activity: An expanded typology. 36(1): 32-72.
Review, 9: 248-258. W. H. 2003. Econometric
Academy of Management Greene, analysis (5th ed.).
Besley, T., & Coate, S. 2003. Elected versus

appointed Englewood Cliffs, NJ: Prentice-Hall.
regulators: Theory and evidence. Working paper, Grier, K. B., Munger, M. C., & Roberts, B. E. 1994. The
Department of Economics, London School of Eco determinants of industry political activity, 1978
nomics. 1986. American Political Science Review, 88: 911
1994. 926.
Boddewyn, J., & Brewer, T. International business
political behavior: New theoretical directions. Acad Hansen, W. L., & Mitchell, N. J. 2000. Disaggregating and
emy ofManagement Review, 19: 119-143.
explaining corporate political activity: Domestic and
Bonardi, J. P. 2004. Global and political strategies in foreign corporations in national politics. American
deregulated industries: The asymmetric behaviors of Political Science Review, 94: 891-903.

Bergh
Heckman, J. J. 1979. Sample selection bias as a specifi Kaufman, A., Englander, E., & Marcus, A. 1993.
Selecting
cation error. Econometrica, 47: 153-161. and organizational structure for implementing issues
Henderson, R., & Cockburn, I. 1994. compe

management: A transaction costs and agency theory
Measuring
tence? Exploring firm effects in pharmaceutical re perspective. In B. Mitnick (Ed.), Corporate political
agency: 148-168. London:
search. 15: 63-84. Sage.
Strategic Management Journal,
Hersch, P., & McDougall, G. 2000. Determinants of auto

Keim, G., & Baysinger, B. 1988. The efficacy of business
mobile PAC contributions to house incumbents: political activity: Competitive considerations in a
context.
Own versus rival effects. Public Choice, 104(3-4): principal-agent Journal of Management,
14: 163-180.
329-343.
Keim, G., & Zeithaml, C. 1986. Corporate political strat

Hill, C. W. 1990. Cooperation, opportunism and the in
visible hand: Implications for transaction cost the egies and legislative decision-making: A review and
ory. Academy ofManagement Review, 15: 500 contingency approach. Academy of Management
Review, 11: 828-843.
513.
Hillman, A., & Hitt, M. 1999. Corporate political strategy Kingdon, J. 1984. Agendas, alternatives and public pol
formulation: A model of approach, participation and icy. Boston: Little, Brown.
strategy decisions. Academy ofManagement Re Lenway, S., & Rehbein, K. 1991. Leaders, followers and
view, 20: 193-214. free riders: An empirical test of variation in corpo

rate political involvement. Academy
Hillman, A., & Keim, G. 1995. International variation in of Manage
ment Journal, 34: 893-905.
the business-government interface: Institutional and
considerations. Lord, M. 2000. Corporate political strategyand legislative
organizational Academy of Man
Review, 20: 193-214. decision making. Rusiness and Society, 39(1): 76-93.
agement
Hillman, A., Sch?ler, D., & Keim, G. 2004. Macher, J., & Henisz, V. 2004. Firm- and
country-level
Corporate
trade-offsand contingencies in the evaluation of for
political activity: A review and research agenda.
Journal ofManagement, 30: 837-857. eign investment: The semi-conductor industry 1994
2002. Organization Science, 15: 537-554.
Hillman, A., Zardkoohi, A., & Bierman, L. 1999.
Corpo
rate political strategies and firm performance: Indi Mahon, J. 1993. Shaping issues/manufacturing agents:
cations of firm-specific benefits from personal ser Corporate political sculpting. In B. Mitnick (Ed.),
vice in theU.S. government. Strategic Management Corporate political agency: 187-212. London: Sage.
Journal, 20: 67-81. G. 1996. The Commission as a
Majone, European regula
tor. In G. Majone (Ed.), Regulating Europe: 61-79.
Holburn, G. L. F., & Vanden Bergh, R. G. 2002. Policy and
New York: Routledge.
process: A game-theoretic framework for the design
of nonmarket strategy.Advances in Strategic Man Makadok, R., & Walker, G. 2000. Identifyinga distinctive
agement, 19: 33-66.
competence: Forecasting ability in the money fund
Holburn G. L. F., & Vanden Bergh, R. G. 2004. Influencing industry. Strategic Management Journal, 21: 853
864.
agencies through pivotal political institutions. Jour
nal of Law, Economics and Organization, 20: 458 Masters, M., & Keim, G. 1985. Determinants of PAC par
483. ticipation among large corporations. Journal of Pol
G. L. F., & Vanden itics, 47: 1158-1173.
Holburn, R. G. 2006. Consumer
Bergh,
capture of regulatory institutions: The diffusion of McCubbins, M. D., Noll, R. G., &
Weingast, B. R. 1987.
public utility consumer advocacy legislation in the Administrative procedures as instruments of politi
United States. Public Choice, 126(1-2): 45-73. cal control. Journal of Law, Economics and Orga
nization, 3: 243-277.
Holburn, G. L. F., & P. T. 2002. Interest
Spiller, group
representation in administrative institutions: The McCubbins, M. D., Noll, R. G., &
Weingast, B. R. 1989.
impact of consumer advocates and elected com Structure and process, politics and policy: Admin
missioners on regulatory policy in the United istrative arrangements and the political control of
States. POWER Working Paper 002, University of agencies. Virginia Law Review, 75: 431-508.
California, Berkeley. McCubbins, M. D., & Schwartz, T. 1984.

Congressional
L. S. 2000. America's electric utilities: oversight overlooked: vs. fire alarms.
Hyman, Past, police patrols
present, and future (7th ed.). Vienna, VA: Public American Journal of Political Science, 24: 165-179.
Utilities Reports. McGahan, A. M., & Porter, M. E. 1997. How much does
Jacobson, R. 1992. The "Austrian" school of strategy. industry matter, really? Strategic Management
18: 15-30.
Academy ofManagement Review, 17: 782-807. Journal,
D. 2003.
Jones, T. 1995. Instrumental stakeholder theory: A syn Mueller, Public choice III.
Cambridge, U.K.:
thesis of ethics and economics. Academy ofMan Cambridge University Press.

agement Review, 20: 404-442.
Munger, M. 1988. On the political participation of the

1228 Academy Journal December
ofManagement
firm in the electoral process: An update. Public Journal of Economies and Management Science, 2:
Choice, 56: 295-298. 3-21.
Niskanen, W. 1971. Bureaucracy and representative Vietor, R. 1994. Contrived competition: Regulation and
government. Aldine. deregulation in America. Boston: Harvard Univer
Chicago:
to Institutional Pro sity Press.

Oliver, C. 1991. Strategic Responses
cesses. Academy of Management Review, 16: 145 Weatherby, J.L. 1971. A note on administrative behavior
179. and public policy. Public Choice, 11: 107-110.
Olson, M. 1965. The logic of collective action. Cam Weber, M. 1947/1922. The theory of social and eco
bridge, U.K.: Cambridge University Press. [A. M.
nomic organization Henderson & T. Parsons,
trans.] New York: Oxford University Press.
North, D. C. 1990. A transaction cost theory of politics.
Journal of Theoretical Politics, 2: 355-367. Weidenbaum, M. L. 2003. Rusiness and government in
the global marketplace. Upper Saddle River, NJ:
Pfeffer,J.1981. Power in organizations. Marshfield, MA:
Pearson Education.
Pitman.
B. R., & Moran, M. J. 1983. Bureaucratic dis
Weingast,
Pfeffer, J. 1992. Managing with power: Politics and in cretion or congressional control? Regulatory policy
fluence in organizations. Cambridge, MA: Harvard
Press.
making by the federal trade commission. Journal of
University Political Economy, 91: 765-800.
Pfeffer,J.,& Salancik, G. R. 1978. The external control of B. 1984. A resource-view of the firm. Strate
Wernerfelt,
New York: & Row.
organizations. Harper
gic Management Journal, 5: 171-180.
Porter, M. 1980. Competitive strategy. New York: Free Williamson, O. 1975. Markets and hierarchies. New
Press.
York: Free Press.
P. S., Lenway, S., & Govekar, M. 1990.
Ring, Management Zardkoohi, A. 1985. On the political participation of the
of the politicalimperative in international business. firm in the electoral process. Southern Economic
Strategic Management Journal, 11: 141-151. 51: 804-817.
Journal,
Rumelt, R. P. 1991. How much does industry matter?
Strategic Management Journal, 12: 167-185. A
Russo, M. 1992. Power plays: Regulation, diversification,
and backward integration in the electric utility in Jean-Philippe Bonardi (jbonardi@ivey.uwo.ca) is an as
dustry. Strategic Management Journal, 13: 13-27. sociate professor at the Richard Ivey School of Business,
Schuler, D. 1996. Corporate political strategy and foreign University of Western Ontario. He received his Ph.D.
from HEC School of Management in Paris. He studies
competition: The case of the steel industry. Acad in regulated
relations and
Journal, 39: 720-737.
business-government deregu
emy ofManagement
lated industries, as well as in the context of international
Sch?ler, D., Rehbein, K, & Cramer, R. 2002. Pursuing trade negotiations.
strategic advantage through political means: A mul
tivariate approach. Academy of Management Jour Guy L. F. Holburn is an assistant professor at theRichard
nal, 45: 659-672. Ivey School of Business, University ofWestern Ontario.
He received his Ph.D. from the Business and Public Pol
Shaffer, B. 1995. Firm-level responses to government reg
ulation: Theoretical and research
icy Program at theHaas School of Business, University of
approaches. Jour His examines
California, Berkeley. research corporate po
nal ofManagement Studies, 21: 495-514.
litical strategy and international investment strategy.
Shaffer, B., Quasney, T., & Grimm, C. 2000. Firm-level
performance implications of nonmarket actions. Richard G. Vanden Bergh is an assistant professor at the
Business and Society, 39(2): 126-143. School of Business Administration, University of Ver
mont. He received his Ph.D. from the Business and Pub
Silverman, B. 1999. Technological resources and the direc
lic Policy Program at the Haas School of Business, Uni
tion of corporate diversification: Toward an integration
of the resource-based view and transaction cost eco versity of California, Berkeley. He studies political
institutions and the nonmarket strategy of firms.
nomics. Science, 45: 1109-1124.
Management
Stigler, G. 1971. The theory of economic regulation. Bell A


Academy of Management

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Academy of Management

Caricato da

Copyright:

Formati disponibili

Nonmarket Strategy Performance: Evidence from U.S.

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

NONMARKET STRATEGY PERFORMANCE:

RICHARD G. VANDEN BERGH

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

political markets more or less attractive from a THEORETICAL BACKGROUND

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

Demanders of public policy Suppliers of public policy

Focal Firm Elected Polticians

Political Markets and the Role of

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

perfect understanding of the issues

icantly higher than the current minimum wage, but

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

oversight of regulatory agencies. so, regulators adhere to the procedural constraints

2. Rivalry between politicians 1989).

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

Nonmarket related to the performance of its nonmarket

from other firms' interactions with government

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

EMPIRICAL INVESTIGATION firms' decisions to file with the U.S. International

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

Sample Initiation = 1 if Att > = 0 otherwise.

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

.10 .05 .19 .03

-.05 .11 -.06.11

.01 .12 .04 .05 .01 .31

.14 .06 .06 .15 .21 .04

.07 .60 .06 .00 .05 .01 .12

10. Cumulative rate reviews .16 Change

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

TABLE 3 the selection effect, we could not interpret the co

Hypothesis 2: Political rivalry suggests that activists constitute a particularly dif

and statistically significant at the 5 percent level, a

Note, however, that, because of the nonlinearity of commission."

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

ticularly well organized interest group, to "pur Our

evidence may reflect the weaker ability of the ap

negative and statistically significant coefficient on Cumulative rate reviews 0.19**

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

variable was whether or not a utility initiated a rate TABLE 6

model variables with PUC budget -0.03

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

Bonardi, J. P., & Keim, G. 2005. Corporate political strat

Conclusion egies forwidely salient issues. Academy ofMan

be done. de Figueiredo, J.M., & Tiller, E. H. 2001. The structure

Besley, T., & Coate, S. 2003. Elected versus

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

Henderson, R., & Cockburn, I. 1994. compe

Hersch, P., & McDougall, G. 2000. Determinants of auto

Keim, G., & Zeithaml, C. 1986. Corporate political strat

view, 20: 193-214. free riders: An empirical test of variation in corpo

California, Berkeley. McCubbins, M. D., & Schwartz, T. 1984.

thesis of ethics and economics. Academy ofMan Cambridge University Press.

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM

to Institutional Pro sity Press.

Stigler, G. 1971. The theory of economic regulation. Bell A

This content downloaded from 140.182.72.234 on Mon, 7 Apr 2014 00:59:38 AM