Organizational structures in Information technology startups

Boundary Spanning Roles and Organization Structure^

HOWARD ALDRICH NYSSILR, Cornell University

DIANE HERKER Oakland University

Boundaries are a defining characteristic of organizations, and bound- ary roles are the link between the environment and the organization. The creation, elaboration, and functions of boundary spanning roles are examined, with attention to environmental and technological sources of variation in the structure of boundary roles. Eleven hypoth- eses integrate the material reviewed and are amenable to empirical test. Future research should overcome problems created when organ- izations are treated as "wholes" or single entities.

A minimal defining characteristic of a formal organization is the distinction between members and non-members, with an organization existing to the extent that some persons are admitted,

Howard Aldrich (Ph.D. — University of Michigan) is Associate Professor of Organizational Behavior at the New York State School of Industrial and Labor Relations, Cornell University.

Diane Herker (Ph.D. candidate — Cornell University) is In- structor in Management, School of Economics and Manage- ment, Oakland University.

Received 6/9/75; Revised 12/22/75; Accepted 6/7/76; Revised 7/13/76.

while others are excluded, thus allowing an ob- server to draw a boundary around the organiza- tion (61, pp. 139-146). Defining organizations in

1 The senior author thanks his colleagues, Sam Bacharach and Robert Stern, for admirably fulfilling their roles as good listeners and trenchant critics. Richard Hall,Jeff Pfeffer, Steve Rosell, and Dave Whetten also contributed to the clarity of the argument. Sandra Miller of the Centre for Environmental Studies, London, provided valuable assistance in the prepa- ration of the paper.

This review could not have been written without the foundation laid by James D. Thompson,and our debt to him should be evident to those familiar with his work.

217

218 Boundary Spanning Roles and Organization Structure

terms of boundaries to interaction also allows a parsimonious definition of the role of formal au- thority in an organization: authorities are per- sons who apply organizational rules in making decisions about entry and expulsion of members (6, p. 283). In this sense, organizational behavior (OB) has always contained an implicit "open sys- tems" view, although few theorists or research- ers have studied boundary-maintaining or boundary-crossing (5).

The definition and location of a specific boundary may be possible only given a specific conceptual and empirical context. This article takes the existence of boundaries as given, while treating boundary spanning activity as problem- atic. Specifically, it examines functions served by boundary roles, the generation of boundary units and roles relating organizations to their en- vironments, and the environmental and organi- zational sources of variation in the structure of boundary roles. Use is made of existing litera- ture, but the argument is speculative at many points.

Although most investigators agree on the importance of focusing on relations between or- ganizations and their environments, there is lit- tle agreement on the degree of autonomy of ac- tion organizations have vis-a-vis their environ- ments. At the extreme are two positions: a na- tural selection model, laying heavy emphasis on the dominance of environmental constraints on behavior, and a strategic choice or resource de- pendence model, emphasizing the active role or- ganizational administrators play in shaping out- comes. As these two macro-theoretical positions are reviewed elsewhere (9), and there is no pros- pect of reconciling them in this paper, we con- centrate on propositions and hypotheses of the middle-range.

Eunctions of Boundary Roles

Two classes of functions are performed by boundary roles: information processing and ex- ternal representation. Information from external sources comes into an organization through

boundary roles, and boundary roles link organi- zational structure to environmental elements, whether by buffering, moderating, or influenc- ing the environment. Any given role can serve either or both functions.

The Infornnation Processing Function

In focusing on the information processing function, we are following the lead of Dill (24), who suggested that the environment of an or- ganization could be treated as information avail- able to the organization through search or expo- sure. Thompson and McEwen discussed the or- ganization's need for information to judge the amount and sources of support for its goals (63, p. 30). Terryberry argued that viable organiza- tions are characterized by "an increase in the ability to learn and to perform according to changing contingencies in the environment" (60, p. 660).

Boundary role incumbents, by virtue of their position, are exposed to large amounts of poten- tially relevant information. The situation would be overburdening if all information originating in the environment required immediate atten- tion. Boundary roles are a main line of organiza- tional defense against information overload (17, 42). Expertise in selecting information is conse- quential, since not all information from the envi- ronment is of equal importance. External infor- mation can be conceptualized in terms of a three-part hierarchy, corresponding to Parsons' (47) distinction between three levels of author- ity in organizations: strategic, managerial, and technical information (17, p. 325). Their relative importance varies by type of environment and technology; e.g., in stable homogeneous envi- ronments and organizations with highly routin- ized technology, strategic information is less im- portant.

The process by which information filters through boundary positions into the organiza- tion must be examined. Boundary roles serve a dual function in information transmittal, acting as both filters and facilitators. Information over-

Academy of Management Review - April 1977 219

load would still be a problem if all relevant in- formation had to be immediately communicated to internal members. Accordingly, boundary role personnel selectively act on relevant informa- tion, filtering information prior to communicat- ing it (23, 40). They act autonomously on some information, and consolidate, delay, or store other information, thus alleviating the problem of overloading communication channels (al- though perhaps incurring other costs to the or- ganization in the process). Information is sum- marized and directed to the organizational units that need it.

Boundary role personnel may act on infor- mation requiring an immediate response, as when a sales department responds to a customer inquiry about product specifications. They may store information for future use, as when a pur- chasing department files information on a new supplier's products, to be referred to at re-order points. Boundary personnel in marketing may uncover trends in the demand for their organi- zation's products which will have a major impact on the mix of resources required, and communi- cate this information to purchasing. Boundary units may also summarize information and com- municate it to other units on a regularized basis (41).

The expertise of boundary role occupants in summarizing and interpreting information may be as important to organizational success as ex- pertise in determining who gets what informa- tion, depending upon the uncertainty in the in- formation processed. Information to be commu- nicated often does not consist of simple verifi- able "facts". If the conditions beyond the boundary are complexly interrelated and cannot be easily quantified, the boundary role incum- bents may engage in "uncertainty absorption", — drawing inferences from perceived facts and passing on only the inferences (40, p. 189).

Consider the case of a lobbyist formulating a report on a bill and amendments that will differ- entially affect the operations of his or her organ- ization. The lobbyist will have to summarize in- formation about the bill's progress, testimony in

hearings, and apparent predispositions of com- mittee members and other legislators, as well as making the entire situation meaningful to his or her superiors. If these superiors cannot under- stand the interrelationships and implications of the raw data, they will not be able to use the in- formation. Some simplification is necessary and the relationships of events in Washington to or- ganizational operations will have to be clearly specified. In short, the lobbyist must put the in- formation in usable form (67).

Innovation and structural change are often alleged to result from information brought into the organization by boundary personnel (11, 28). All complex organizations have a tendency to move toward an internal state of compatibility and compromise between units and individuals within the organization, with a resultant isola- tion from important external influences (18). This trend can jeopardize the effectiveness and per- haps the survival of the organization, unless the organization is effectively linked to the environ- ment through active boundary personnel. By scanning the environment for new technological developments, innovations in organizational de- sign, relevant trends in related fields, etc., boundary personnel can prevent organizations from becoming prematurely ossified and mis- matched with their environments (20).

This review of the information processing function of boundary roles may be summarized in the following hypothesis:

HI : An organization's ability to adapt to envi- ronmental contingencies depends in part on the expertise of boundary role incum- bents in selecting, transmitting, and inter- preting information originating in the en- vironment.

The External Representation Function

External representation can be viewed in terms of an organization's response to environ- mental influence. Environmental constraints and contingencies can be adapted to in at least three ways: (a) by internal structural differentiation to

220 Jilt

Boundary Spanning Roles and Organization Structure f

match the pattern of the relevant environment, which requires information about environmental characteristics; (b) by gaining power over rele- vant elements of the environment, manipulating it to conform to the organization's needs; and (c) a compromise position,the modal pattern of use of boundary personnel in "normaT'bound- ary spanning roles. Included under the external representation function are all boundary roles that involve resource acquisition and disposal, political legitimacy and hegemony, and a resi- dual category of social legitimacy and organiza- tional image.

Boundary roles concerned with resource ac- quisition and disposal include purchasing agents and buyers, marketing and sales representatives, personnel recruiters, admissions officers, and shipping and receiving agents. In these roles the organization is represented to the environment, because the normal flow of authoritative com- mands is from the core of the organization to these boundary roles. The behavior of person- nel in these roles is supposed to reflect the pol- icy decisions of decision makers in line roles.

This usual flow of directives to boundary roles presents two problems for boundary per- sonnel. First, much of the information they at- tend to has an external origin, and it occasional- ly becomes apparent that policy directives are based on information that is no longer relevant. This poses a dilemma for the conscientious boundary spanner, especially in organizations with a high degree of decentralization — should behavior be immediately modified to corre- spond to latest developments, or should action be delayed until the information has "gone through channels?" Second, as Strauss (59) pointed out in his study of purchasing agents, some boundary personnel are not satisfied with their subordinate position on the vertical axis of the organization, given their self-evident hori- zontal location of equality with other depart- ments. Thus dissatisfied boundary spanners take the initiative to increase their power vis-a-vis other units. For example, personnel officers sug- gest changes in job descriptions before agreeing

to post them; admissions or intake staff develop their own criteria of "worthy" applicants (14); and purchasing agents make informal compacts with salespersons from outside firms to push products which their production department "really" needs (59).

Boundary roles involved with maintaining or improving the political legitimacy or hege- mony of the organization not only represent the organization but also mediate between it and important outside organizations. The term "me- diate" refers here to aspects of the boundary role involving negotiations that will eventually affect the power of the focal organization vis-a- vis another organization or group. Kochan (36) notes that city governments have created collec- tive bargaining units as a response to threats to the city's control over its employees. The role of the corporate lawyer is perhaps the most clear example of the necessity and difficulty of pre- serving at least an equal balance of rights and re- sponsibilities between business organizations (38).

Boundary spanning personnel can help maintain the legitimacy of the organization by providing information to important client groups, specially adapted for them. Aldrich and Reiss (10) note that police officers on the beat transmit an image of city law enforcement capa- bilities to small businesspersons independent of the businessperson's attitudes toward the police themselves. Information transmittal is facilitated because both police and small business are ex- posed to environmental forces that make their commonality of interest highly salient. Adair's (1) study of the use of Navaho Indians as health aides for their own communities found that the Indians functioned as mediators in their bound- ary roles, drawing the doctor and the Navaho pa- tient closer together. The Indian health worker offers a different side to each party involved, finding a middle ground to settle discords be- tween them. Detached school workers perform the same sort of representation function for school systems.

Maintaining the organizational image and

>Academy of Management Review - April 1977 221

enhancing its social legitimacy are less a matter of mediating contacts than of simply making the organization visible. Advertising and public re- lations specialists try to influence the behavior of target groups in ways that benefit the focal or- ganization, without bargaining or negotiating with the target group. The flow of intra-organi- zational influence to these roles is much more one-sided than in the boundary roles described above; one apparent consequence is a high rate of turnover.

One function of boards of directors and public advisory commissions is to link the organ- ization to target groups in the environment in a highly visible way, so that they will feel their in- terests are being represented. Thus, women and blacks are being appointed in increasing num- bers to corporate boards, and students now serve as trustees on university boards. Fulfilling this function requires recruiting people who are already members of or in contact with specific target groups. Maniha and Perrow (39) describe the formation of a Youth Commission's board in terms of such community interest groups as Catholics, the university, and youth-oriented voluntary associations.

These three varieties of external represen- tation functions can be related to organizational effectiveness in the second hypothesis:

H2: An organization's ability to cope with en- vironmental constraints depends in part on the ability of boundary role incum- bents to achieve a compromise between organizational policy and environmental constraints, to choose strategic moves to overcome constraints, or to create condi- tions in which the organization's autono- my is seldom challenged.

Creation of Boundary Roles

By definition all organizations have some boundary spanning roles, if only at the level of the organization head or chief executive. But some have an elaborate set of boundary roles

while others have only a few. In some cases boundary roles are formalized into full-time or- ganizational positions, while in others they are only part-time activities. This section examines the generation and formalization of boundary roles as explicit organizational roles, with refer- ences to organizational size and technology, and various characteristics of organizational environ- ments. To understand the process of boundary spanning behavior, an interactive model of the kind developed by Adams (2) is needed, but such models are highly specific to the particular pair-wise relationship being examined. Here we are concerned with the general features of the boundary role while recognizing that actual be- havior in boundary roles will vary from context to context.

The extent to which organizational posi- tions involve interaction with external elements varies greatly. Many positions outside the tech- nical core involve some extra-organizational in- teraction, but only a few involve intensive inter- action. As an empirical test of arguments in this article we would need a measure providing bet- ter than a "yes-no" categorization of roles as either boundary spanning or not. This would re- quire determining the proportion of time spent with outsiders, the number of outside contacts, the importance of each contact, etc., as Whetten (65) has recently done in a study of manpower organizations.

The number of formally designated bound- ary spanning roles in an organization is partially dependent on organizational size (5). A small or- ganization is able to survive with a fairly simple structure, with relatively few differentiated roles and functions (16, 21, 22). Its structure may be less formalized and more amenable to restruc- turing to achieve and maintain a satisfactory po- sition vis-a-vis its environment. A small organiza- tion might be willing and able to rely on infor- mation brought to it informally by its members. This tendency is more marked among organiza- tions that have highly committed members or that are not highly dependent on their environ- ment for survival (6), such as a small religious sect

222 Boundary Spanning Roles and Organization Structure

(29). As organizational and environmental com- plexities increase, organizations can no longer afford non-differentiated boundary spanning activities.

Technology and Boundary Role Differentiation

Holding size constant, boundary spanning units or roles should be expected to increase as a proportion of all roles as organizations differ- entiate in response to the interaction of technol- ogy and environment, and under the direct im- pact of environmental pressure. In the following discussion, technology is treated as a source of internal differentiation generating boundary roles to the extent that varying technology types create different patterns of organization-envi- ronment interaction. Thompson's categories of mediating, long-linked and intensive technology capture the implications of various technology types for the generation of boundary roles (62, pp. 15-18).

Organizations with a mediating technology link clients or customers with each other, as in the case of banks, insurance companies, or the post office; or they link clients with other organ- izations, thus serving a "people-processing func- tion" (30). Such organizations should have the highest proportion of boundary roles, as boundary roles are their line roles. The whole- saler of small consumer goods has boundary personnel who purchase goods from producers and sell them to organizations which, in turn, sell them to retail customers. An investment banking firm contracts with a client to put to- gether a "package" of investment instruments that satisfy financial needs and then sells the "package" to other organizations with funds to invest. Boundary personnel similarly serve a line function in the people processing component of organizations such as schools and government agencies.

A study of organizations using a mediating technology in the book publishing and record producing industries found that they allocated a large proportion of their personnel to boundary

spanning roles, on the input side to contract for and supervise the production of raw material, and on the output side to promote the cultural products and achieve optimal distribution (32). These boundary roles also monitored the envi- ronment and provided information quickly to managers and executives, apparently as a strate- gy to help the organization hold its position in a very uncertain environment.

Organizations with a long-linked technol- ogy attempt to buffer most of their units and roles from the environment, and have a lower proportion of boundary roles. Since the various organizational units are serially interdependent, there are many boundary roles between intra- organizational components, but the focus here is on roles at the external boundary. Specific boundary roles are important for such organiza- tions. First, long-linked technology gains maxi- mum efficiency through standardized produc- tion of large volumes of output (to take advan- tage of economies of scale) and so such organi- zations need an effective marketing and sales force (35, 51). Second, Thompson (62, p. 40) ar- gues that "organizations employing long-linked technologies and subject to rationality norms seek to expand their domains through vertical integration", and thus the legal and accounting departments of such organizations interact fairly intensively with potential acquisitions in the en- vironment (50).

Organizations using an intensive technol- ogy, which depends on the object being worked on, also buffer most of their roles from the envi- ronment. They often achieve this by temporarily drawing the object or the client into the organi- zation. In intensive technology organizations concerned with people-changing activities, the client is temporarily assigned an organizational role, and must change behavior to suit norms which preclude appealing to his or her environ- mental role relative to the organization. The boundary personnel who engage in initial inter- action with potential clients affect the organiza- tion's subsequent internal operations if they have the power to admit or reject clients, e.g..

Academy of Management Review - April 1977 223

the physician associated with a hospital or the admissions officer at a private college. By detect- ing a violation and making an arrest, a police of- ficer provides the rest of the criminal justice sys- tem with raw material to be processed (55).

Thompson argues that these organizations seek to expand their domains by incorporating the object worked on, with "total institutions" (26) placing an almost impenetrable boundary around clients. The people-changing organiza- tions that use an intensive technology (e.g., hos- pitals) have one characteristic that opens them to environmental influence — their high degree of professionalization. Aiken and Hage (4) assert that professionals in organizations enage in a great deal of boundary-spanning contact be- cause of the need to maintain contact with a pro- fessional reference group and keep abreast of changing technology in their field.

In intensive technology organizations not concerned with people-changing activities, the clients often become a temporary part of the or- ganization's administrative structure and thus need a liaison person to represent them, as in the construction industry (62, p. 44). Thus, while it is clear that mediating technology organiza- tions have proportionately more boundary span- ning roles than other organizations, the relative ranking of organizations with long-linked and intensive technology cannot be determined without further empirical research.

H3: Organizations using a mediating technol- ogy will have the highest proportion of boundary roles, while organizations using long-linked and intensive technologies will have a smaller proportion of bound- ary roles.

H4: Organizations using long-linked or inten- sive technologies will departmentalize and otherwise separate boundary span- ning units from their core technical units.

Environment and Boundary Role Differentiation

Environmental pressures are responsible for much of the observed differentiation In organi-

zations, after technology is taken into account. Some theories of organization-environment in- teraction posit that maintenance of a high de- gree of internal organizational complexity oc- curs only in response to environmental pres- sures that tolerate nothing less (9, 18). The con- centration of important environmental elements into an organized form may promote a matching organizational response, in the form of more boundary units or more formalized and central- ized boundary spanning activities (52). Kochan points out that:

A number of collective bargaining researchers have noted the proliferation of specialized la- bor relations units in city governments in re- sponse to the increased unionization of pub- lic ennployees (36, p. 7).

Wilson (67) discusses the growth of lobbying ef- forts of unions, trade associations, and other or- ganizations representing vested interests, in re- sponse to the growth in power of the federal government. The consumer, ecology, and other movements have brought pressure on corpora- tions, which have responded by establishing public relations units to deal with pressure groups. The same type of response occurred among public agencies in the President's Office for Consumer Affairs and similar offices in HEW, HUD, etc. (45).

H5: Organizations in environments where im- portant elements are concentrated will have a higher proportion of boundary roles than organizations in environments where important elements are dispersed.

Heterogeneous environments should evoke more organizational boundary spanning units and roles, as organizations "seek to identify homogeneous segments and establish structural units to deal with each" (62, p. 70). Separate units, whether established on the basis of heterogene- ity in a client population or in the geographical domain served, lead to a higher proportion of boundary roles than in organizations of compar- able size serving a homogeneous domain. Hos- pitals establish separate units for obstetrics, con-

224 Boundary Spanning Roles and Organization Structure

tagious diseases, and out-patient services. Auto manufacturing firms respond to heterogeneity in client income distribution by divisionalizing op- erations around products with different selling prices, but not necessarily costs (19).

H6: Organizations in heterogeneous environ- ments will have a higher proportion of boundary roles than organizations in homogeneous environments.

Stable environments, which presumably call for less frequent monitoring, should evoke fewer boundary roles than unstable environments, al- though much depends on whether change is oc- curring at a constant or variable rate (62, pp. 71- 72). In the cultural industry (books, records, films) where fashions change rapidly, we would expect to find proliferation of boundary roles on both input and output sides of the organization (32). In organizations producing for a stable mar- ket, we would expect most roles to be related di- rectly to the production process, although an un- expected shift in the market can change the sit- uation dramatically, as in Emery and Trist's (25) example of a food canning firm. Some theorists have argued that the most salient characteristic of organizational environments today is their rate of change (60), a purported trend which should cause an increase in the proportion of boundary spanning roles in most organizations.

H7: Organizations in rapidly changing envi- ronments will have a higher proportion of boundary roles than organizations in sta- ble environments.

A final dimension to be considered is the ex- tent to which the environment is rich or lean in resources (7, 12, 25). In rich environments, hold- ing competition constant, we would expect to find fewer boundary roles, since environmental search and monitoring would be less critical for organizational survival than in environments where lack of resources prevents the accumula- tion of a "resource cushion". At a time of inter- national or interorganizational hostility, the en- vironment becomes less rich in information and

so nations and organizations have to allocate more roles to their boundaries to make use of what little infornnation is available (66).

H8: Organizations in lean environments will have a higher proportion of boundary roles than organizations in rich environ- ments.

Environment and Boundary Role Formalization

If internal complexity is associated with en- vironmental pressures and demands, organiza- tional boundary roles will be officially recog- nized full-time roles, especially if decision mak- ers recognize the existence of such contingen- cies. Whether boundary roles will be thus for- malized depends upon organizational recogni- tion of potentially costly contingencies that may arise from failure to maintain effective links to elements in the environment. But such recogni- tion need not be based on intelligence that or- ganization itself has accumulated, as profession- al education (e.g., MBA programs), professional and trade publications (e.g.. Business Week), and informal inter-firm contact (51) all keep organi- zational decision-makers abreast of new devel- opments in the design and administration of for- mal organizations (12, 58). The following discus- sion focuses on direct, rather than indirect, rec- ognition of environmental contingencies and constraints.

Most large organizations formally designate such roles as labor negotiators and corporate lawyers responsible for transactions in the labor relations sector, since strikes and law suits might cripple an organization. Labor contracts are ne- gotiated for fairly long periods of time and the organizational costs of mistakes in boundary- spanning negotiation with unions are fairly high. Smigel's (57) discussion of staff recruitment in large Wall Street law firms indicates that firms be- came aware of a variety of changes in their envi- ronment, including the small output of prestigi- ous law schools, students' wariness of accepting positions with large firms, and the increasing de- mand for trained lawyers. One result was crea-

Academy of Management Review - April 1977 225

tion of the formal position of "hiring partner" to scan the potential output of top ranking law schools, sell students on advantages of employ- ment in the firm, and thus improve recruitment of desirable graduates.

The more critical the contingency, the more attention is paid to explicit formalization of the role and selection of an incumbent. This is par- ticularly evident with regard to the composition of boards of directors of large organizations, as Pfeffer argues that organizations:

use their boards of directors as vehicles through which they co-opt, or partially ab- sorb, important external organizations with which they are interdependent (50, p. 222).

Price (53), in a study of state wildlife governing boards, found that one major function of board members was to serve as a buffer group between the full-time staff and the public. Zald points to the external representation functions of boards of directors:

They promote and represent the organiza- tion to major elements of the organizational set, for example, customers, suppliers, stock- holders, interested agencies of the state, and the like. That is, they defend and support the growth, autonomy, and effectiveness of their agencies vis-a-vis the outside world (68, p. 99).

Another critical contingency for large cor- porations involves managing reciprocal relations with other large firms; a trade relations person is alerted to look for opportunities to cooperate with other firms when it could be to their mutual advantage. Perrow notes that:

the practice of reciprocity is so extensive that about 60 percent of the top 500 corporations have staff members who are explicitly assigned to trade relations. Of course, any smart sales executive or top executive can serve in this capacity. However, it is striking that the prac- tice is sufficiently common to justify so many special positions among the giants (48, p. 122).

Pursuing leads on possible acquisition of other companies is an important function as- signed to corporate development units. Aguiler notes that the high volume of acquisition leads generated by this staff:

demonstrates how the formalization of a search procedure can significantly increase a company's relative involvement with a partic- ular kind of information (3, p. 47).

H9: Boundary roles are most likely to be for- malized when crucial environmental con- tingencies have been explicitly recognized by organizational decision makers, or the organization is structured in a way that fa- cilitates the adoption of structural innova- tions through imitation and borrowing from other organizations or other external sources.

Routinization, Discretion, and Power

The degree of role specificity (31) of bound- ary roles varies considerably, with some being highly routinized and others highly non-routine. Thompson (61) identified two conditions leading organizations to increase specificity of control over boundary role personnel. First, organiza- tions that provide services for large numbers of persons and thus face many non-members (rela- tive to members) at the boundaries of the organ- ization must either substantially increase the number of personnel in boundary positions or else routinize the tasks of existing staff so they can handle a higher volume of work. Second, organizations using a mechanized production technology which places a premium on large runs of standardized products depend upon a large volume of standardized transactions per member at the organization's output boundary. Pressures for routinization are somewhat les- sened when the non-members dealt with have little or no discretion to participate in a relation- ship. Later Thompson identified a third condi- tion, in that stable environments are likely to produce boundary roles governed by rules, whereas unstable environments are likely to in- crease flexibility in the specificity of boundary role routines (62, p. 71).

Purchasing agents and sales personnel inter- act frequently with suppliers and buyers and usually deal with fairly homogeneous groups of

226 Boundary Spanning Roles and Organization Structure

organizations and individuals. A high frequency of interaction and homogeneity of elements at the boundary allows behavior in these roles to be highly routinized (27, p. 24; 62, p. 111). Rou- tinization is reflected in the existence of standard purchase and sales forms or contracts, standard operating procedures for soliciting and accept- ing bids, and standard operating procedures for calling on customers and closing sales. A classic example is the retail salesperson who knows the one proper way to record a cash sale and the one proper way to record a credit transaction.

Routinization of roles at the organization's boundary not only increases efficiency in han- dling predictable relationships and large num- bers of repetitive transactions, but also serves a social control function. The programmed nature of these activities is partial insurance of boundary spanner consistency with organizational pro- cedures, norms, and goals. Members who inter- act freely with non-member groups, particularly homogeneous sets, are likely to develop attitudes consistent with those of the non-members, rath- er than of their focal organization. The existence of standard operating procedures partially pro- tects the organization against attitudes and be- haviors that are not consistent with organization- al objectives.

MacAuley (38) noted different behavioral orientations among boundary and non-bound- ary personnel in the use of contracts among busi- ness firms. Sales departments tended to display non-organizational norms that made them will- ing to conduct transactions without legally binding contracts. Members of the controllers and legal departments upheld the organizational norm of using contracts, which also happened to be the norm of their professional reference group.

Mathiesen's (41) study of prison staff mem- bers identified the boundary role of "social worker" as a position that was difficult to routin- ize.

Though almost all staff members claimed there were few or no specific rules or regulations guiding their communications, the social

workers appeared particularly vehement about it, and included relations to official or- ganizations. They stressed that here they had to be extremely flexible; that they had to or- ganize the work on a day-to-day basis and ac- cording to the unique circumstances of the individual case (41, p. 76).

Telephoning was preferred to the use of letters, and when complex cases arose, face-to-face meetings were arranged. Boundary roles (once created) that deal with heterogeneous elements must contain a minimal degree of routinization to maximize flexibility in dealing with special cases.

The degree to which boundary roles are routinized thus is a function of both the need to adapt to environment contingencies and con- straints, and the need to control behavior of po- tentially deviant members. Routinization can serve as a social control mechanism when the organization does not or cannot assume norma- tive commitment of members to organizational procedures. Similar mechanisms would be the use of uniforms to reinforce organizational identification, or frequently shifting employees between boundary roles and core roles to pre- vent development of identification with ele- ments in the environment (27, p. 21).

H10: Boundary role routinization will vary di- rectly with the volume of repetitive work, the predictability of outcomes, the homo- geneity and stability of the environment, and the need to control the behavior of organizational members.

Power in Boundary Roles

Thompson (62) noted that where the envi- ronment is heterogeneous and shifting and where contingencies are important to the or- ganization, boundary personnel are expected to exercise discretion and develop expertise, and to the extent they are successful in recognizing contingencies, they may become powerful with- in the organization. The potential power posi- tion of boundary spanners was evident in discus- sion of their infornriation-processing function.

,Academy of Management Review - April 1977 227

The information that filters into the organization through boundary positions is often not raw data, but the inferences of boundary role in- cumbents.This type of information is difficult for anyone removed from the boundary to verify. The process of uncertainty absorption is a case of creation of organizational intelligence; and once created, intelligence tends to be accepted (66).

The organization thus relies upon the ex- pertise and discretion of its boundary role per- sonnel. They have a gatekeeper's power, and may become even more powerful if they make correct inferences and if the information is vital for organizational survival (40). Their power is further enhanced to the degree that the nature of the task assigned the boundary role makes routinization of the role difficult, if not impos- sible.

Pettigrew (49, p. 190) showed how the "self- interested filtering of information during a deci- sion process by a gatekeeper" enabled a bound- ary spanner to consolidate and enhance his or her power. To the extent that information access and control is a power resource, boundary span- ners are in an excellent structural position to convert this resource into actual power.

Labor negotiators provide an example of a boundary role that is difficult to routinize, thus leading to de facto concentration of power in the role. Even though negotiators may deal with fairly homogeneous groups, the outcomes are not highly predictable and the costs to the or- ganization may be high. Therefore, negotiators require some degree of discretionary power. Their power is enhanced to the extent that the group they are negotiating with is powerful.

Specifically, a number of components of un- ion power that derive from the tactics or ac- tivities of the union - involvement in city elec- tions, use of strike threats in bargaining - all are associated with a higher degree of power in the boundary unit in city governments (34, p. 27).

H11: The power of boundary role incumbents will vary inversely with boundary role rou-

tinization, and directly with their own ex- pertise in accomplishing role require- ments and with the costliness and unpre- dictability of interorganizational transac- tions.

Organizational dependence on boundary role personnel raises the issue of their commit- ment to and integration into the organization. The least costly monitoring mechanism is for the organization to rely on the professional identifi- cation and ethics of the boundary personnel. More obtrusive strategies include attempts to in- doctrinate-boundary personnel in organizational policies, norms, and goals, prior to their engag- ing in interorganizational contacts. Rotation of members among boundary roles is another ac- tive strategy, although it has costs in terms of disrupting local adaptations that have been made by boundary spanners. An organization might grant powerful boundary personnel higher posi- tions within the organizational structure to rein- force commitment, although such positions may be a result of the power these members develop through successful interaction on behalf of the organization, e.g. the common practice of pick- ing top management out of the sales division of an organization.

Many studies emphasize the stress and con- flict felt by personnel in boundary roles (33, 46), but overlook the positive potential inherent in their role accumulation prospects. Sieber (56) has recently argued that multiple relationships with diverse role partners provide numerous sources of gratification, rather than strain, to individuals such as boundary personnel. He notes that role rights and privileges may accumulate more rapid- ly than duties, that overall status security may be enhanced by means of buffer roles, that multiple roles can serve as resources for status enhance- ment and role performance, and finally that mul- tiple roles may enrich one's personality and en- hance one's self-conception. While this article is not addressed to the issue of costs and benefits to individuals who occupy boundary roles, the positive side of boundary spanning activities

228 Boundary Spanning Roles and Organization Structure

should be seen as a counter to the negative im- age currently portrayed in the literature.

Two recent studies in a research and devel- opment organization and a large manufacturing company report positive correlations between boundary spanning activity and several dimen- sions of job satisfaction (34, 35). These studies al- so found very small or insignificant correlations between role conflict, role ambiguity, and boundary spanning activity. The authors argue that boundary spanning jobs, to the extent they enable role incumbents to reduce uncertainties for others, permit boundary spanners to gain power, improve their bargaining position, and hence increase their job satisfaction and perhaps even gain better jobs.

Implications

The picture of boundary spanning roles por- trayed in this article has two implications for the study of formal organizations. First, this view of organization-environment interaction is a de- cidedly disaggregated one, in contrast to current literature which sees organizations responding as "wholes" to environmental influence (15, 25). We treat boundary spanning roles as the critical link between environmental characteristics and

organizational structure, with the further stipu- lation that organizations face multiple environ- ments and thus can have a variety of boundary roles of units with different structural character- istics. This implies, for example, that when an in- vestigator studies the impact of interorganiza- tional dependence on organizational structure, the place to begin is with its impact on bound- ary spanning roles in the immediate vicinity of the dependence relationship, rather than with the structure of the organization as a whole (8, 43).

Second, more empirical studies are needed of how personnel in boundary spanning units or roles carry out their duties, and in particular how such role performance varies under different en- vironmental conditions and over time. This would mean more studies of the type carried out by Mintzberg (44), on the day-to-day be- havior of managers, or Mathiesen (41) on the day-to-day behavior of staff members in two Scandinavian prisons. Both studies make exten- sive use of non-participant observation and de- tailed first-hand knowledge of the actual, rather than self-reported, behavior of those persons studied. The cumulation of such studies would enable us to understand the process by which boundary spanning roles are generated, elab- orated, and used by their incumbents.

REFERENCES

3.

4.

Adair, Ross. "The Indian Health Worker," Human Or- ganization, Vol. 19 (1960), 59-63. Adams, J. Stacy. "The Structure and Dynamics of Behav- ior in Organizational Boundary Roles," in M. Dunette (Ed.), Handbook of Organizational and Industrial Psy- chology (Chicago: Rand McNally, 1976), pp. 1175-1199. Aguiler, Francis. Scanning the Business Environment (New York: MacMillan, 1967). Aiken, Michael, and Jerald Hage. "Organizational Inter- dependence and Intra-organizational Structure," Amer- ican Sociological Review, Vol. 33 (1968), 912-930.

5. Aiken, Michael, and Jerald Hage. "Organizational Per- meability." Paper presented at the 1972 Meetings of the American Sociological Association.

6. Aldrich, Howard. "Organizational Boundaries and Inter- organizational Conflict," Human Relations, Vol. 24 (1971), 279-287.

7. Aldrich, Howard. "An Organization-Environment Per- spective on Cooperation and Conflict in the Manpower Training System," in A. Negandhi (Ed.), Conflict and Power in Complex Organizations (Kent, Ohio: C.A.R.I., Kent State University, 1972), pp. 11-37.

Academy of Management Review - April 1977 229

8. Aldrich, Howard. "Organization Sets, Action Sets, and Networks: Making the Most of Simplicity," in P. Nystrom and W. Starbuck (Ed.), Handbook of Organizational De- sign (Amsterdam: Elsevier. 1977).

9. Aldrich, Howard, and Jeffrey Pfeffer. "Environment of Organizations," in A. Inkeles (Ed.), Annual Review of Sociology, Vol. II (Palo Alto, Calif.: Annual Review Inc., 1976), pp. 79-105.

10. Aldrich, Howard, and Albert ). Reiss, Jr. "Police Officers as Boundary Personnel," in H. Hahn (Ed.), Police in Ur- ban Society (Beverly Hills, Calif.: Sage Publications, 1971), pp. 193-208.

11. Bennis, Warren. Beyond Bureaucracy (New York: Mc- Graw-Hill, 1966).

12. Benson, |. Kenneth. "Models of Structure Selection." Unpublished paper. University of Missouri-Columbia, 1971.

13. Benson,). Kenneth. "The Interorganizational Network as a Political Economy," Adm/nistrat/ve Science Quarterly, Vol. 20 (1975), 229-249.

14. Blau, Peter. The Dynamics of Bureaucracy (Chicago: Uni- versity of Chicago Press, 1957).

15. Blau, Peter, and Richard Schoenherr. The Structure of Organizations (New York: Basic Books, 1971).

16. Blau, Peter, and W. Richard Scott, formal Organizations (San Francisco: Chandler, 1%2).

17. Brown, W. B. "Systems, Boundaries, and Information Flow," Academy of Management journal. Vol. 9 (1966), 318-327.

18. Campbell, Donald. "Variation and Selective Retention in Socio-Cultural Evolution," General Systems, Vol. 16 (1969), 69-85.

19. Chandler, Alfred. Strategy and Structure, (Cambridge: MIT Press, 1962).

20. Child, John. The Business Enterprise in Modern Indus- trial Society, (London: Collier MacMillan, 1969).

21. Child, John. "Parkinson's Progress: Accounting for the Number of Specialists in Organizations," Administrative Science Quarterly, Vol. 18 (1973), 328-348.

22. Child, John. "Participation, Organization and Social Co- hesion," Human Relations, forthcoming.

23. Cyert, Richard and James March. A Behavioral Theory of the firm (Englewood Cliffs, N.J.: Prentice-Hall, 1963).

24. Dill, William. "The Impact of Environment on Organiza- tional Development," in S. Mailick and E. Van Ness (Eds.), Concepts and Issues in Administrative Behavior (Engle- wood Cliffs, N.J.: Prentice-Hall, 1962).

25. Emery, Fred, and Eric Trist. "The Causal Texture of Or- ganizational Environments," Human Relations, Vol. 18 (1965), 21-31.

26. Goffman, Erving. " O n the Characteristics of Total Institu- tions," in D. Cressey (Ed.), The Prison (New York: Holt, Rinehart, and Winston, 1961).

27. Guetzkow, Harold. "Relations Among Organizations," in R. Bowers (Ed.), Studies on Behavior in Organizations (Athens, Georgia: University of Georgia Press, 1966).

28. Hage, Jerald, and Michael Aiken. Socia/ Change in Com- plex Organizations (New York: Random House, 1970).

29. Harrison, Michael. "The Adjustment of a Social Move- ment to its Organizational Environment." Unpublished paper. Department of Sociology, SUNY, Stony Brook, 1972.

30. Hasenfeld, Yeheskel. "People Processing Organizations: An Exchange Approach," American Sociological Review, Vol. 37 (1972), 256-263.

31. Hickson, David. "A Convergence in Organization The- ory," Administrative Science Quarterly,Vol. 2 (1966), 224-237.

32. Hirsch, Paul. "Processing Fads and Fashions: An Organi- zational Set Analysis of Cultural Industry Systems," /Amer- ican yourna/o/Soc/o/ogy Vol. 77 (1972), 639-659.

33. Katz, Daniel, and Robert Kahn. The Social Psychology of Organizations (New York: John Wiley and Sons, 1966).

34. Keller, Robert, and Winford Holland. "Boundary-Span- ning Roles in a Research and Development Organiza- tion: An Empirical Investigation,"/Academy o/Manage- ment journal. Vol. 18 (1975), 388-393.

35. Keller, Robert, Andrew Szilagyi, and Winford Holland. "Boundary-Spanning Activity and Employee Reactions: An Empirical Study," Human Relations, forthcoming.

36. Kochan, Thomas. "Determinants of the Power of Bound- ary Units in an Interorganizational Bargaining Relation," /\dm/n;s(raf/ve Science Quarterly, Vol. 20 (1975), 434-452.

37. Lawrence, Paul, and Jay Lorsch. Organization and Envi- ronment (Cambridge, Mass.: Graduate School of Busi- ness Administration, Harvard University, 1967).

38. MacAulay, Stewart."Non-contractual Relations in Busi- ness: A Preliminary Study," American Sociological Re- view, Vol. 28 (1963), 55-67.

39. Maniha, John, and Charles Perrow. "The Reluctant Or- ganization and the Aggressive Environment," Adminis- trative Science Quarterly, Vol. 10 (1965), 238-257.

40. March, James, and Herbert Simon. Organizations (New York: John Wiley and Sons, 1958).

41. Mathieson, Thomas. Across the Boundaries of Organiza- tions (Calif: Glendessary Press, 1972).

42. Meier, Richard. "Information Input Overload," in F. Massarik and P. Ratoosh (Eds.), Mathematical Explora- tions in Behavioral Sciences (Homewood, III.: Richard Irwin, Inc., 1965).

43. Mindl in, Sergio, and Howard Aldrich. "Interorganiza- tional Dependence: A Review of the Concept and a Re- examination of the Findings of the Aston Group," Ad- ministrative Science Quarterly, Vol. 20 (1975), 382-392.

44. Mintzberg, Henry. The Nature of Managerial Work (New York: Harper and Row, 1973).

45. Nadel, Mark. The Politics of Consumer Protection (In- dianapolis: Bobbs-Merrill, 1971).

230 Boundary Spanning Roles and Organization Structure

46. Organ, Dennis. "Linking Pins Between Organizations and Environment," Bus/ness Horizons, Vol. 14 (Decem- ber 1971), 73-80.

47. Parsons, Talcott. "Suggestions for a Sociological Ap- proach to Theory of Organizations," Administrative Sci- ence Quarterly, Vol. 1 (1956), 63-69.

48. Perrow, Charles. Organizational Analysis (Belmont, Calif.: Wadsworth Publishing Company, 1970).

49. Pettigrew, Andrew. "Information Control as a Power Resource," Sociology, Vol. 6 (1972), 187-204.

50. Pfeffer, Jeffrey. "Size and Composition of Corporate Boards of Directors," Administrative Science Quarterly, Vol. 17 (1972), 218-228.

51. Pfeffer, Jeffrey, and Huseyin Leblebici. "Executive Re- cruitment and the Development of Interfirm Organiza- t ion," Administrative Science Quarterly, Vol. 18 (1973), 449-461.

52. Pfeffer, Jeffrey, and Huseyin Leblebici. "The Effect of Competition on Some Dimensions of Organizational Structure," Social forces. Vol. 52 (1973), 268-279.

53. Price, James. "The Impact of Governing Boards on Or- ganizational Effectiveness and Morale," Administrative Science Quarterly, Vol. 8 (1963), 361-368.

54. Pruden, H., and R. Reese. "Interorganizational Role-Set Relations and the Performance and Satisfaction of Indus- trial Salesmen," Administrative Science Quarterly, Vol. 17 (1972), 601-609.

55. Reiss, Albert J. The Police and the Public (New Haven: Yale University Press, 1971).

56. Sieber, Sam. "Toward a Theory of Role Accumulation," American Sociological Review, Vol. 39 (1974), 567-578.

57. Smigel, Erwin. "The Impact of Recruitment on the Large Law Firm," Human Relations, Vol. 25 (1960), 56-65.

58. Starbuck, William. "Organizations and Their Environ- ments," in M. Dunnette (Ed.), Handbook of Organiza- tional and Industrial Psychology (Chicago: Rand Mc- Nally, 1976), pp. 1069-1124.

59. Strauss, George. "Tactics of Laterial Relationship: The Purchasing Agent," Administrative Science Quarterly, Vol. 7 (1962), 161-186.

60. Terryberry, Shirley. "The Evolution of the Organizational Environments," Administrative Science Quarterly, Vol. 12 (1968), 590-614.

61. Thompson, James. "Organizations and Output Transac- tions," American journal of Sociology, Vol. 68 (1962), 309- 325.

62. Thompson, lames. Organizations in Action (New York: McGraw-Hill 1967).

63. Thompson, James, and William McEwen. "Organization- al Goals and Environment," American Sociological Re- view, Vol. 23 (1958), 23-31.

64. Weber, Max. The Theory of Social and Economic Organ- ization (New York: The Free Press, 1947).

65. Whetten, David. Predicting Organization-Set Dimen- sions: An Inter-organizational Study of the Effectiveness of Manpower Programs in New York State (Ph.D. Thesis, Cornell University, 1974).

66. Wilensky, Harold. Organizational Intelligence (New York: Basic Books, 1967).

67. Wilson, James Q. Political Organizations (New York: Basic Books, 1973).

68. Zald, Mayer. "The Power and Functions of Boards of Di- rectors," American journal of Sociology, Vol. 75 (1%9), 97-111.