The Blackwell Companion to Sociology (78 page)

development laboratories typically lacked funds, public legitimacy, and in-

house capability to market their products and maneuver through the govern-

mental regulatory labyrinth. They formed alliances with resource-rich diversified corporations that could provide badly needed financial infusions. In return, the large firms sought to acquire tacit knowledge and to learn new technological

skills from their junior partners. In a study of equity joint ventures among US, Japanese, and European new materials, industrial automation, and automotive

products firms, Gulati (1995b) found evidence consistent with both resource

dependence and social structural explanations. Strategically interdependent

companies (firms operating in complementary market niches) forged more alli-

ances than did firms with comparable resources and capabilities. Previously

allied partners had a higher probability of forming new alliances with one

another, suggesting that `òver time, each firm acquires more information and

builds greater confidence in the partnering firm'' (p. 644). But additional alliances eventually reduced the probability of forming further agreements because

of fears of lost autonomy by becoming overly dependent on a single partner.

Indirect connections through the prior alliance network also shaped succeeding

alliances with new partners. That is, previously unconnected firms were more

Networks and Organizations

333

likely to collaborate if they both had ties to a common third party. Gulati

concluded that ``the social network of indirect ties is an effective referral

mechanism for bringing firms together and that dense co-location in an alliance

network enhances mutual confidence as firms become aware of the possible

negative reputational consequences of their own or others' opportunistic beha-

vior'' (p. 644).

Interorganizational

Interorganizational Networks and Political Action

Political action by interorganizational networks is the primary theme of the

organizational state theory, developed in Laumann and Knoke's (1987) research

on US energy and health policy-making and elaborated in a comparative

project on US, German and Japanese labor policy networks (Knoke et al.,

1996). The core actors in the organizational state are formal organizations,

both governmental and non-governmental, that attempt to realize their public

policy interests through collective political action. Corporations pursue pro-

tected markets and greater profits, labor unions seek higher wages and better

working conditions, professional associations try to protect their members'

autonomy, political parties promote ideologies and solicit votes for re-election, and bureaucracies seek to enforce agency mandates and boost budgets. The main

outcomes of organizational state processes are collectively binding decisions in specific policy events, such as legislative acts, regulatory decrees, court rulings, or strong bureaucratic leadership.

The policy domain is the largest unit of analysis in organizational state theory.

It is defined as a social system within which collectively binding decisions are made, implemented, and evaluated with regard to a specific type of public policy.

A domain is comprised of all the important actors holding common interests in

those policies, regardless of whether they agree on preferred outcomes to policy events. Indeed, every policy domain is split among two or more opposing blocks

of political organizations attempting to persuade policy-making authorities to

approve different options. The alternative that eventually prevails results from more or less intense political struggles to assemble sufficient support to pass a bill or enact a regulation. Policy domains develop fairly stable power structures dominated by the most powerful peak interest groups and governmental actors.

Gaining access to a domain's central positions requires an organization to

acquire and deploy information (both technical expertise and political know-

ledge) and resources (both material and symbolic).

A central task for policy domain analysts is to uncover the key structural

relations in the political networks among the core organizations. The two

especially relevant networks are information exchange and resource networks.

These relations most closely parallel Knoke's (1990, pp. 11±16) analytic distinction between two basic power dimensions: `ìnfluence,'' persuasive communica-

tions intended to change others' beliefs and perceptions regarding political

actions, and ``domination,'' resource transactions of physical benefits (or

harms) in return for compliance with commands. Information is an intangible

334

David Knoke

asset ± such as scientific, legal, or political knowledge ± whose transmission from one actor to another does not result in its loss to the first possessor. Resources are physical commodities ± such as money, labor power, and facilities ± whose

control can be transmitted from one actor to another.

A fundamental organizational state proposition is that both information and

resource networks shape organizational efforts to focus sufficient power on

other policy domain organizations, most importantly on particular public

authorities responsible for ultimately deciding a policy event, to sway the col-

lective decision in their favor. Organizations on both sides of a policy conflict seek to increase the probability that these authorities will choose the outcome

they prefer. To a great extent, actors use pre-existing information and resource exchange networks in their influence efforts on specific policy events. Hence,

organizational positions in the already existing information and resource

exchange networks are important factors both in stimulating participation and

in determining the outcomes of policy events.

During the course of public policy struggles, opposing coalitions or alliances

among organizations typically coalesce to undertake coordinated political

actions aimed at furthering their members' common interests. A collective action involves three or more organizations working together in an effort to obtain

their preferred policy event outcome. The necessary preconditions that lead to

collective action involve mutual recognition by organizations that they share

common goals, followed by formation of linkages that enable them to undertake

cooperative activity. Some collective actions are aimed at supporters or potential supporters, while others are targeted toward neutral observers (e.g. the media

and the general public) and against active opponents (more often to dissuade

them from taking opposing action than to convert them). But the most important

targets in any policy domain are the governmental officials who possess the legal or customary authority to make a decision about an event that is binding on the

domain as a whole.

Network analysis reveals that global patterns of information exchange among

political organizations in a policy domain are structured according to their

common policy interests. Very dense communication networks typically connect

a domain's core organizations. In each of the three US domains more than one

hundred trade associations, labor unions, professional societies, public interest groups, federal executive agencies, and congressional committees were asked

with which organizations they shared (i.e. both sent and received) important

policy information (Laumann and Knoke, 1987; Knoke et al., 1996). The density

of communication ties was 0.38 for the US labor domain and 0.30 in both the

energy and health domains. Diagrams representing the communication network

structures in the three domains showed that organizations expressing similar

issue interests and policy preferences tended to occupy the same regions. In a

two-dimensional plot of the labor policy organizations' path distances (i.e. the number of direct or indirect steps required to connect a pair of organizations), most unions, business associations, and federal agencies were located in three

distinct sectors (Knoke et al., 1996, p. 112). Similarly, sharp cleavages occurred within the energy policy domain between consumers and several types of energy

Networks and Organizations

335

producers, and in the health policy domain between consumers, clients, medical

researchers, and health care providers (Laumann and Knoke, 1987, pp. 242,

246). This partitioning into specialized policy segments reflects a tendency for political organizations to communicate primarily with their friends and potential allies, and to avoid giving information to organizations with which they share

few common interests.

Every domain's communication structure also revealed that governmental

actors and interest groups with broad policy agendas filled the central locations.

Thus, the center of the energy domain was occupied by the White House,

Department of Energy, and key House and Senate energy and resources commit-

tees. This core group was closely surrounded by such major labor unions and

trade associations as the AFL-CIO, United Automobile Workers, American

Petroleum Institute, Edison Electric Institute, American Gas Association, Amer-

ican Mining Congress, and National Automobile Dealers Association (Laumann

and Knoke, 1987, pp. 243±5). Given the divergent and sometimes antagonistic

policy interests of these organizations, their close proximity to the center of the domain's communication network indicates that these major political actors

maintained high levels of information exchange with potential opponents.

Open communication carries risks as well as benefits: `Àt a minimum, it can

alert them that a policy change is afoot. It also conveys one's position, and it could convey extensive information about the political environment. As a strategic blunder, direct contact can provide the impetus for latent opposition to

mobilize, expanding the scope of conflict. As a tactical matter, it is clearly in a group's interest to let its opponents address the information problems on their

own'' (Hula, 1999, p. 54). To lower the risks of aiding potential foes, political organizations confine their exchanges of sensitive tactical and strategic information only to other actors sharing their preferred policy outcomes. Communica-

tions with policy opponents are more likely to serve as warnings and deterrents, signaling the intensity of organizational interests and intentions to fight on

particular policy events.

Exchanging policy-relevant information with other political organizations is

an important prerequisite for engaging in political action, but is insufficient to explain organizational attempts to influence policy-makers' decisions. Engaging

in policy discussions is a relatively low-cost activity, where participants can

quickly move into or drop out of contact with one another. The complexity

and density of interorganizational communication networks makes their bound-

aries amorphous and renders large, diffuse discussion groups unwieldy as effect-

ive policy influencing instruments. More promising insights accrue by examining

the formation of advocacy coalitions whose member organizations cooperatively

pursue collective policy goals. The expanding population of Washington polit-

ical organizations and the complexities of gaining access to federal institutions noted above increasingly necessitate coalitional behavior. Organizations can

leverage their political influence by pooling resources, especially their technical and political expertise, to create an effective division of labor for contacting governmental targets. Public officials may be more easily impressed, even convinced, by arguments advanced through a broadly united front than by clashing

336

David Knoke

claims proffered by individual organizations. Hence, the main incentives for

joining a collective action are reducing resource costs while increasing the

chances of achieving a successful political outcome.

Organizations that form coalitions to pool their political resources are gen-

erally more successful in realizing their goals than are actors attempting by

themselves to affect policies. Because considerable resource expenditures are

involved, only organizations holding high interests in a particular policy event typically pursue their preferences collaboratively with others. The configurations of allied and opposing actors may be quite fluid and complex when examined

over time. Further, because events reflect numerous decision points within a

continuing policy stream, the interweaving of actors, relations, and actions is

best considered across multiple, sequential policy events. A large sample of

events can reveal both stability and change in policy network structures as well as the impact of event outcomes on subsequent policy struggles.

Comparative labor policy domains research found significant network effects

on both organizational influence reputations and policy event activity (Knoke et al., 1996). The more central an organization in both the communication network (measured by policy information exchanges) and the support network

(measured by resource exchanges), the higher its reputation as an especially

influential player in labor policy. Similarly, greater centrality in both networks leads to more involvement across numerous legislative events in six types of

political influence activities, including coalitions with other organizations. In the US and German cases, communication centrality exerted a much stronger effect