Computers and computer networks as catalysts of intraorganizational decentralization
Institute of Sociology at the University
Institute of Sociology at the University of Zurich
Studying the organizational implications of new
communication technologies is a highly complex enterprise:
In contrast to mere "terminals" as they existed in
the seventies, personal computers are highly polyvalent tools potentially able
to accomplish an unlimited variety of different tasks with an even larger
variety of alternative procedures.
When computers are internetted, the spectrum of alternatives widens again immensely, because users get access to external information and can interact, exchange data and cooperate in any possible ways. In a more general perspective, several studies support the contention that the implementation of IT-technologies is a conflictive political process shaped by various individuals and groups that try to maximize their own preferences and goals (Newman and Noble 1990; Bloomfield and Daniel 1995; Sabherwal and King 1992).
Some authors see the antecedent organizational structures as the main or even sole determinant. so that the results of IT implementation processes are predetermined by the formal power structure existing at the time of inception (Franz and Robey 1986; Ang and Cummings 1997; Tractinsky and Jarvenpaa 1995). Others give more room for technological influences and unplanned contingencies by maintaining that IT-implementations set into motion various adaptation processes from which new cooperation patterns and even power structures may finally emerge (Lee 1991; Tan et. al. 1995; William and Wilson 1997). Rob Kling agrees with Bullen and Bennett (1990) that in contrast to classical machinery, installing the hardware is not sufficient for making the technology running and reaping its fruits. Instead, long processes of software development (and learning how to use them effectively) are necessary for exploiting the potentials of stand alone computers (Kling and Iacono, 1989), and even more encompassing processes of social reorganization are indispensable for maximizing the returns of Intranets, Extranets, groupware programs and systems of multilateral communication.
As a consequence, it is not to be expected that structural effects (like shifts in the distribution of power and influence) will actualize as soon as larger number of employees sit behind the computer. Instead, they may slowly become visible to the degree that management and workers discover how the new technologies can be instrumentalized for fulfilling their needs and furthering the goals of their particular organization.
2. The issue of (de-)centralization
In the sphere of conventional industrial technology, it is evident that many centralizing effects stem from the basic fact that management controls its implementation. Thus, especially Neomarxist authors like Braverman (1974) have stressed that material technology functions as a medium for perfectionizing interclass domination, because capitalists can control worker behavior know all details by developing and applying machineries that predetermine the bodily movement, the pace of work as well as the pattern of interpersonal cooperation.
It is argued that in the same way, management also controls the implementation of computer technologies by deciding about the hardware to be bought and the software to be implemented. As in the classical epochs of industrialization, private enterprises have been the major innovators: by investing into huge new computerized technologies that have affected the single workplace and work team as well as the functioning of large-scale factories and interorganizational systems. Consequently, it is the workplace where individuals are most deeply affected by the new computer technologies, so that their roles and behavioral patterns have often been fundamentally transformed:
By increasing the visibility and measurability of individual performances, many of these applications evidently increase the range of centralized human control:
Countering the optimistic visions of electronic democratization with a Neomarxist perspective, Rob Kling maintains that office automation follows the same logic of interclass domination as the classical industrial mechanization:
In a similar vein, Wresch argues that centralizing effects dominate because the management has the power of specifying the ways the new technologies are implemented: e. g. by choosing the type of network administration software to be used. In accordance with their basic control interests, they tend to use it for installing a perfectionized "panopticon": a device of surveillance that allows them to track remotely all movements and activities of their employees (Wresch 1996). However, it is doubtful whether this entails the same specific discretion as in the case of classical machinery, because the implemented technologies are usually open for highly unpredictable and polyvalent usage.
First of all, even stand-alone computers may empower individuals in many different ways: by extending their range for competent individual actions and decisions, even when their level of formal qualification is quite low.
Secondly, the instalment of a computer network goes along with many risks because it provides a multitude of additional communication channels that make more unpredictable who communicates when with whom about what topic - and therefore also: who is in possession of what kind of information at a specific point of time. To make such specifics predictable and controllable would be equivalent of closing the system down. A fortiori, opening the access to the Internet amplifies the spectrum of possible usage patterns: e. g. by facilitating the rapid switching between intra- and extraorganizational roles. (Geser 2002) 
All these potentials can only be exploited to the degree that no codified rules and no centralized controls determine who has to use the computer at moment in what precise way: so that employees are free to choose optimal procedures in the light of given goals, needs, and circumstances. As a consequence, it becomes more difficult to prescribe exact procedures and to base work coordination on behavioral control. Instead, management has to shift to output control: by assessing only work results, regardless of the way by which they have been achieved Baloh/Trkman 2003). This again implies that the main motivation of employees is to solve tasks and to reach goals, not to conform to bureaucratic norms and procedures.  (Baloh/Trkman 2003) More than that: a certain degree of ex ante decentralization and informality may even be necessary for effective IT exploitation: an environment where employees have enough autonomy and leeway to learn on the job, and to try many risky things out without pressure to be effective and without fear of sanctions.
All these arguments accord with the
finding of Hitt and Brynjolfsson 1997 that high levels of IT investment and IT
use) were related to works system with decentralized authority. Among the earliest proponents of this optimistic
view, Lee Sproull and Sara Kiesler (1991) assert that computer-mediated
communication promotes more equalized power structures by allowing peripheral
employees to circumvent hierarchical barriers and to participate more actively
in the processes of intraorganizational communication. Additional equalization
effects may stem from the fact that receivers are often not informed about the
sender’s status characteristics (gender, education or formal position).
Eric Lee has found that such empowerment of lower levels takes place for different reasons in various parts of the organizations. In the administrative sector, it is mainly stemming from increased centrality in communication networks, while in the technical sector; it is predominantly associated with increased resource provision (Lee 1991).
On a most general level, computer mediated communication (CMC) reduces the role of hierarchical structures because it increases the capacity of horizontal interaction networks to exchange information, to coordinate work processes and to reach various decisions (Bishop/ Levine 1999). In particular, CMC has the advantage that such processes can be started informally by anybody without much costs and efforts, so that the initiating or catalyzing role of formalized groups and organizations (e. g. unions) is sharply reduced. (Bishop/Levine 1999) Given their capacity to support any kind of decentralized communication, Intranets are intrinsically apt to facilitate the emergence of grassroot Networks of workers and to promote processes of widespread horizontal coordination and collectivization (Levine 1999). While almost unknown in 1993, Intranets expanded in following years so quickly that in 1996, 64% of the Fortune 1000 companies had them already installed (Sinton 1996). By providing bilateral and multilateral communication channels across the organization irrespective of IT-specificities at different locations, Intranets are highly polyvalent technologies promoting the downward flow of information, facilitating the universal access to central documents as well as supporting and kind of horizontal interactions and group activities on all organizational levels (Marchand et. al 2001). 
In fact, by strengthening the action potentials of horizontal networks, IT is threatening the traditional hierarchical structures that are primarily based on vertical interactions. For instance, the interaction barriers between different departments may collapse in a way that the authority of middle managers is effectively undermined (Martin 2002). In addition, online interaction may well function as a breeding ground for a new corporate culture that is primarily based on interpersonal goodwill and trust (Nandhakumar, 1999; Martin 2002). Online based networks can in turn increase the company's cohesion by binding the employees tighter into the organization (Finholt, Sproull, and Kiesler 1990).
In addition, Intranets may also promote corporate democracy by catalyzing vertical upward communication. Employees may feel more encouraged to articulate their views directly to top managers because they can avoid the tensions usually associated with face-to-face communication (Martin 2002). In his case study on a large high technology enterprise, Bishop and Levine found that CMC facilitated upward communication, thus enhancing the employee’s ability to express their views and to participate in decision-making processes (Bishop/Levine 1999).
On the one hand, increased upward flows of communication may enhance productivity by accelerating decision making and providing managers with better information (Ichniowski et al. 1996), on the other hand, they can also improve the worker's job satisfaction and their overall evaluation of the firm (Lawler et al. 1995). However, such direct vertical connections may also have countervailing centralizing effects because peak managers have more chances to make their influence felt on the lower organizational levels by simply circumventing all intermediate ranks. Thus, some more sophisticated studies have led to the paradoxical conclusion that IT technologies may foster centralization and decentralization at the same time.
On the one hand, centralization on the encompassing level of the whole organization is increased, because top management uses the new technologies for reinforcing overall formalization and control. On the other hand, this same overall centralization opens the way for delegating more authority to lower levels, because even supervisors with very high discretion can now still be effectively controlled. (Dawson/McLoughlin 1986; Dean et al. 1992; Zeffane 1989)
3. The crucial role of intervening variables
All the straightforward bivariate hypotheses mentioned above collide with the evident fact that technology-structure relationships are mediated by powerful intervening variables.
Thus, Susan Herring has concluded that the technology itself had no intrinsic democratizing potential because personal status factors like gender were decisive whether electronic upward communication was effective or not:
Likewise, Mantovani (1994) concluded that while CMC may overcome geographic obstacles, it does not eliminate social barriers: so that the effects of new Net technologies vary according to the culture, goals and structural conditions of the particular organization where they are applied.
One of the most encompassing research projects was the 10 year study conducted by Symons who examined virtual departments in various organizational settings of firms with 3,000 to 5,000 full-time employees. He came to the conclusion that under certain conditions, new communication technologies can even enhance the power of managers by increasing their capacity to monitor employees. (Symons 1997) On the basis of such arguments, Nerissa Nelson concludes that
Because IT technologies are so polyvalent, any organization can use them to realize its preferred structures to a higher degree. Thus, voluntary associations may use them to reach better approximation to their democratic power profile, while authoritarian bureaucracies may implement them for expanding he scope of hierarchical control (Geser 2001).
In line with this argumentation, Carter has found that IT contributes to decentralization in “organic” organizations that face unstandardized tasks and dynamic environmental change, while they reinforce centralization in the case of “mechanical” systems confronted with more stable conditions (Carter 1984). In addition, it has been observed that the type of work decides whether and to what degree IT is used for individual control. In administrative work flows, optimal conditions exist for applying behavioral and output control at the same time: In the realm of sales however, only output controls are usually applied, and in knowledge work, no control procedures seem viable at all (Frenkel et al. 1999).
We may hypothesize that all occupational role incumbents wish to instrumentalize the new information technologies for furthering their personal and role-related needs - but that they succeed very unevenly according to their highly divergent autonomy, power and material resources. Thus, architects or designers use computers to rationalize the elaboration of new constructions, journalists need the Internet for a quicker scanning of relevant information; and professional scientists make use of mail lists for improving their worldwide communication:
While professionals may thus enhance the scope and variety of their work, lower occupations may find themselves more constrained because they have to tolerate the infringements caused by the expansion of higher occupations. Thus, secretaries lose classical spheres of competence because their bosses use their own computer for mailing messages and their own cell phone for answering calls (Geser/Bürgisser 1998), and workers face perfectionized systems of surveillance because managers seek their self-realization mainly in the implementation of electronic systems of control. As a consequence, we may expect that when most employees are highly qualified professionals, computers and computer networks contribute to decentralization: because members are eager as well as capable of using the new potentials for empowering purposes. On the other hand, organizations staffed with highly unqualified personnel may experience centralizing impacts, because average workers are not able to resist the implementation of managerial systems of taylorist standardization, surveillance and control. In turn, computerization may reinforce deskilling because roles are simplified to a degree that highly unqualified personnel can be hired:
In a similar vein, it was found that the impact of information technologies on hierarchical structure was conditioned by the antecedent degree of centralization. In highly centralized firms, top managers used the new technologies for reducing the middle management levels, in decentralized settings; they instrumentalized them for increasing these middle layers (Pinsonneault and Kraemer, 1997)
The technological impacts may also vary according to organizational size. In his early comparative study of newspaper organizations, Carter concludes that the organizational power structure is more affected by IT in smaller than in larger firms (Carter 1984). This may be explained by the higher structural formalization and inertia of larger enterprises or vice versa: by the higher structural informality and “malleability” of many (especially newer) small firms. On the other hand, Zeffane has found in Australia that in medium to large firms, computer use was significantly related to decentralization (Zeffane 1989). Evidently, the advantages stemming from the loosening of spatial restrictions accrue to mainly larger organization where members with common interests have no chance to meet face-to-face (Bishop/Levine 1999).
Apart from the many conditioning factors on the organizational levels, the microsocial impacts of information technologies may also be co-determined by macrocultural, political and legal-institutional conditions. In his broad survey of European countries, Eric Lee has found that most Nations have provisions to protect the private sphere of employees at the workplace, so that employers have only a limited permission to supervise and control the Email traffic and the Internet surfing activities of their staff (Lee 2000). In addition, Art. 28 of the "European Social Charta" implies that Labour Unions must be given access to the intrafirm computer networks to the degree that such access is necessary for them to gather important information and to contact and organize their members (Lee 2000). In Switzerland, Art. 179 of the penal code disallow employers to supervise private employee activities at the work place, except in cases where such controls are necessary for reasons of security or for assessing the work performance. In addition to such legal provisions, the federal deputy for data security has enacted a guideline (“Leitfaden”) which regulates neatly the conditions under which such supervisions are permitted. In the United States, however, privacy claims at the work place have usually not been protected by the courts (Mansfield/Gabel 2003). Generally, employers are allowed to supervise any online activities of their workers as long as the electronic equipment used is owned by the firm. Therefore, tendencies toward centralized surveillance have become very widespread, especially among larger corporations:
However, such surveillance efforts are often self-defeating because morale is lowered and trust relations are impaired (Schwartz 2000; Mansfield/Gabel 2003).
4. Aims of the study
Based on a comprehensive comparative survey of enterprises from the industrial and the service sector in Switzerland, the following empirical study tries to clarify whether and how computer usage is affecting various organizational conditions at the workplace level, especially the role and status of lower-level operative employees. Given the extensive sample as well as the large variety of measured variables, we are able to address several questions that have remained open by the rather sketchy and inconclusive research results available at this present time.
First of all, we acknowledge that the terms "computer technology" and "information technology" are highly synthetic concepts including many facets that deserve to be distinguished on the analytical as well as on the empirical plane In a sociological perspective, it is particularly crucial to make a difference between
There are good reasons to assume that each higher level of network integration is associated with additional pushes for decentralization, because the spectrum of alternative actions available to any users (even on lowest hierarchical levels) is immensely increased.
Secondly, we take
into account that terms like "centralization" or "decentralization" are also
multidimensional concepts encompassing very different aspects that may not covary highly among each other.
Of course, a full picture would have to include the whole profile of power and
influence over all hierarchical levels. so that basic relationships between
workers and superiors are considered as well as the distribution of competences
branch managers and the CEO.
It is a major limitation of our study that only the lowest level (affecting the
singular workplace of nonsupervisory employees) is taken into consideration. As
a consequence, no countervailing effects on different hierarchical levels can be
detected. On the other hand,
the multidimensionality of the centralization concept is respected by taking
into account that supervisory authority (or its mirror: employee empowerment)
may be different in various aspects of the work process.
Third, we are impressed by the insight that technological impacts on social structures (if such causalities exist at all) are likely to need time, so that they may not easily be discovered in simultaneous cross-sectional correlations. As a consequence, we add some additional empirical evidence based on processes of recent change, not on the current state of conditions.
Fourth, we take the contention serious that computer technology impacts may not be all pervasive, because so many intervening variables determine whether centralizing are decentralizing consequences prevail. Given that our data set extend to almost 2000 cases, we have ample opportunities to test such assumptions by multivariate statistical procedures. In particular, it will be checked whether effects vary across economic branches or according to firm size and the skill level of employees.
And fifth, finally, we hypothesize that decentralization effects on the organizational level go along with correlates on the level of individual roles. In particular, the assumption is tested that technologically induced worker empowerment is associated with a broader range of work tasks and with an increased emphasis on personal self-direction.
5. Data and Methodology
empirical findings are based on two comprehensive survey studies conducted 1998
and 2000 in Switzerland. In both studies, a questionnaire was mailed out to
about 6000 Swiss Firms of all major branches in the industrial and service
sector, out of which about 1500-1800 were finally returned. The sample was
constituted by the KOF (Konjunkturforschungsstelle der ETH Zürich) with the
explicit aim to mirror the Swiss private economy as whole: excluding the
educational health and social welfare sector. For approaching this goal, a
stratified selection was made for generating an equilibrated sample of different
branches on the one hand and different size categories on the other. Special
efforts were made to receive a positive response from the small group of largest
enterprises. In the survey of 1998, the main
focus was laid on staff composition and the demand for occupational skills as
well as on strategies of firm reorganization and ongoing processes of
organizational change. In 2000, the questionnaire addressed usage patterns of
computers and computer network technologies as well as various aspects of
organizational structure. In both survey, the main focus was directed to the
work places of average non-supervisory employees. For measuring the independent variables, informants
of the second survey (= general firm managers) were asked about the percentage
of work places which had access to a computers or to various forms of online
communication. On the level of
the total sample, the following frequency distribution has been achieved:
Table 1: Distribution of firms according to the percentage of work places having access to computers and online communication: total sample (percentage values).
6. Empirical Results
6.1 The variety of work
shown in Table 2, the extended usage of Personal Computers goes along with more
challenging work roles that contain a richer variety of different tasks,
especially among smaller firms with 100 or less employees). It might be
hypothesized that the labour division within smaller enterprises is fluid enough
to be easily affected by technological conditions, while in larger firms; role
specialization is sufficiently fixed by formalized norms and procedures.
Table 2: Variety of work tasks* of a typical employee and the percentage of work places having access to PC: according to size of firm.
* Five point scale between 100
(very low) to 500 (very high)
From Table 3, it can be seen that task variety is
only marginally increased when computers are also internetted. Only in the
category of smaller enterprises (between 41 and 100 employees, is significant
relationship with access to intraorganizational networks can be observed.
Table 3: Correlations between variety of work tasks and percentage of work places with access to various information technologies: according to the size of firm.
* p < .05 ** p < .01
It might be assumed that such job enrichment effects only occur in enterprises where employees are skilled enough to make use of the new facilities, while in branches with unskilled personnel, information technologies may even contrarily be used to increase the degree of specialization. This hypothesis is neatly borne out by Table 4 where we can see that positive correlations only appear when the percentage of unskilled staff is extremely low. Interestingly, the effects emanating from access to the WWW are much stronger than those associated with internal or external electronic mail. Thus, we may conclude that full Internet access is contributing significantly to the richness of work roles, but only when a certain skill level (implying a sufficient “Net literacy”) is given.
Table 4: Correlations between variety of work tasks and percentage of work places with access to various information technologies: according to the percentage of unskilled employees.
* p < .05 ** p < .01
6.2 Distribution of power between workers and supervisors
In a second step, we ask how the introrganizational usage of information technology is related to the centralization of authority on lower (operative) levels of organizations. This balance was assessed by asking informants to characterize the relative influence of workers and first-level supervisors on various work aspects by using a five point scale ranging between 100 (complete autonomy of the worker) to 500 (complete determination by the supervisor).
As shown in Table 5, supervisory iurisdiction is
consistently highest in the small minority of firms that still don’t make any
use of computers, and consistently at the lowest level in the most progressive
firms where more than 80% of all workplaces have digital equipment. Even more
impressive is the faultless monotonic relationship between the percentage of PC-
equipped collaborators and all aspects of supervisory power. Interestingly, the
most pronounced gain in worker empowerment takes place in the external
interaction with clients, while internal task distribution is somewhat less
affected by technological conditions. It could be argued that such correlations are
spurious because they conceal underlying branch-specific factors. For instance,
they may be explained by the regularity that older mass production industries
are centralized as well as “computer-resistent”, while modern high-tech branches
have implemented more progressive organizational styles characterized by
extended computer usage on the one hand and participative employee relations on
Table 5: Authority of supervisors in different domains of work organization*: according to percentage of work places with PC (all firms).
The large sample size evidently provides the opportunity for detailed breakdowns in order to assess whether such decentralization effects are particular to specific industries, or whether they occur as generalized effects across all economic branches.
Table 6 demonstrates convincingly that the second
alternative holds. Out of 95 correlation coefficients, only 11 show a positive
sign and none of them on as significant level. On the other hand, 22 of the
remaining 84 coefficients reach significance, most of hem on the level of one
Nevertheless, industry seems to be more affected than the service sector, where
only firms specialized on wholesale trade show consistently high correlations. Of course, these findings provide no information
whether decentralization results simply from work stations as stand alone units,
from computer networks within the firm (e. g. intranets) or from the integration
of computers into the worldwide Email system and the WWW. In order to disentangle these
different effects, the partialized impacts stemming from access to intrafirm
email, external Email and to the WWW have also been empirically assessed. Surprisingly, the results of Table 7 indicate again
that most effects are originating from stand-alone computers, not from computer
networks, because most negative bivariate correlations with the extensity of
Email and WWW access are reduced to zero when the total number of PC’s is
controlled. Only the control over work specification becomes somewhat more
decentralized when widespread online communication (especially intrafirm Email
Table 6: Correlations between total authority of supervisors in different domains of work organization and the percentage of work places with access to PC in 19 branches of industry and the service sector.
Table 7: Correlations between total authority of supervisors in different aspects of work organization and percentage of work places with access to various information technologies: bivariate and partial correlations.
* p < .05 ** p< .01
Considering the conflicting theorizing and empirical evidence about the relationship between technological decentralization effects and organizational size, the correlations of Table 8 indicate that both sides may be right because a curvilinear relationship stands out. While effects in very small firms are considerable, they are lowest in middle-sized establishments with 101 to 200 employees, and they rise above this level to reach maximum values in firms larger than 500. Especially the control of workers over task distribution and client relations is considerably increased in larger computerized plants.
Table 8: Correlations between the authority of supervisors in different domains of work organization and the percentage of work places with PC: according to the size of firm.
* p < .05 ** p< .01
More detailed analyses (not presented here) support
the conclusion that in all size categories, all major decentralization effects
are exerted by the mere presence of computers, not by their integration into
intrafirm networks or the global Internet. Given the regularity that computers are
significantly more widespread in firms with highly qualified staff, it might be
argued that it is the skill level rather than the technology that is causally
related to the lower level of centralization. Of course, also mutual
contingencies between the two potential determinants could be envisaged. Thus,
it could be hypothesized that decentralization effects of digital technologies
cannot be realized on low skill levels, because employees lack the qualification
for using them effectively for enhancing their own discretion. This would imply
that computers have a “prismatic” effect: by amplifying the polar difference
between rather centralized low-skill and highly decentralized high-skill
organizations. In order to get insight into such complexities, we analyze how the correlations
vary across different levels of skill.
Table 9: Correlations between the authority of supervisors in different domains of work organization and the percentage of work places with PC: according to the percentage of unskilled personnel.
* p < .05 ** p< .01
6.3 Participation in reorganization processes
In order to tap
more extensive organizational empowerment effects transcending the microlevel
interaction with immediate supervisors, it was asked whether management was
formally consulting the nonsupervisory employees when reorganization measures
were planned and implemented. It was found that in exactly 50% of all cases,
some mode of formal consultation was observed. Ranging from simply informing
workers about ongoing reorganization processes to giving them a substantive say
in the respective decisions.
As seen in Table 10, percentages rise consistently with the frequency of
computerized workplaces: achieving maximum values in larger firms with a
penetration rate of more than 80%. Again, we observe that correlation
coefficients covary with firm size in a curvilinear fashion: reaching
(nonsignificant) minimal values within the range between 100 and 500 employees. Further statistical procedures (not presented here)
lead once more to the conclusion that the major empowerment effects stem from
workstations as standalone devices, because no additional empowerment effects
are observed when computers are wired to intraorganizational networks or to the
Table 10: Percentage of firms where nonsupervisory employees are consulted in processes of reorganization: according to the percentage of work places having access to a PC: by size of firm.
Again, we may hypothesize that only
more educated workers may profit from such effects, because they are better able
to use computer for amplifying their level of information and judgment. However,
the results do not allow such a conclusion To the contrary, highest percentages
of worker participation are found on fully computerized enterprises with a very
large percentage of unskilled employees. (Table 11).
Table 11: Percentage of firms where nonsupervisory employees are consulted in processes of reorganization: according to the percentage of work places having access to a PC: by percentage of unskilled employees.
6.4 Recent trends of decentralization
Cross sectional comparative studies provide only limited insight into causalities because organizational effects of computer technology need much time to realize. In contrast to most industrial machineries which enforce specific roles and cooperative patterns at the very moment they are set into motion, computers are just universal tools that display their consequences to the degree their potentialities are discovered and exploited. Such developments imply cumbersome and time-consuming learning processes on all scales: on the organizational level where strategies of computerization and informatization have to be designed and implemented in accordance with the firm’s specific needs and conditions; and on the level of individual employees who have to become acquainted with software systems and to develop usage patterns in the course of daily work experience. While longitudinal studies would be necessary to scan such developments, we try to grasp part of the indirectly by asking whether computer usage has gone along with correlative changes in organizational structure. Thus, while full-blown traditional bureaucracies may still not have transformed into informal “adhocracies” even when everybody works with digital equipment, they may still be show moderate recent trends toward less centralization. For assessing such tendencies, informants were asked whether in the preceding years, reorganization measures aiming at
have been enacted.
While it is well known that under
the concepts of “lean production”, “management reengineering” or various other
headings, such reorganizations have started in the eighties well without being
conditioned by computers or any other technological factors (see Kling 1996),
they may well be reinforced by computers and computer-supported networks,
insofar as capacities for individual discretion as well as for horizontal
interindividual coordination are increased. In accordance with such
argumentations, it is found that the inclination to implement decentralizing
measures is consistently at a minimum in firms where almost no computers are
used, and they reach a peak level when more than 40% of workers use digital
equipment. Interestingly, the impact on interindividual cooperation patterns
(=group works) is stronger than the effects on than the autonomy or voice of
Table 12: Percentage of firms that have experienced recently various changes toward organizational decentralization: according to the percentage of work places having access to a PC.
Given the semantic similarity of the four items as well as their parallel covariation with technology, it seems justified to aggregate them to a summative index that measures the firm’s general tendency toward structural decentralization. As shown in Table 13, only smaller and middle-sized firms seem to be affected, because all correlations vanish above a threshold of 200 employees. This certainly disaccords with the strong correlations between computer usage and current decentralization levels found in larger firms (see Table 8 and Table 10). However, this inconsistency should induce too much speculation, because the sample for which decentralization indices are available is rather reduced.
Table 13: Correlations between the index of recent decentralization1) and percentage of work places with access to various information technologies: according to the size of firms.
1) Sum of the four recent changes mentioned in table 9: ranging between -4 centralization on all aspects) to +4 (decentralization on all aspects)
In addition, Table 13 makes again
evident that all correlations with the Net variables are spurious, because they
are reduced to insignificance when the number of PC’s is controlled. This result
is certainly surprising in the light of all arguments that see the computer
networks, not stand alone computers, as the major agent of organizational
change. Once again, no clear corroboration is
found for the contention that higher levels of vocational education are
preconditions for technological effects of decentralization, because significant
correlations are found on various levels of skill (Table 14).
Table 14: Correlations between the index of recent decentralization1) and percentage of work places with access to various information technologies: according to the percentage of unskilled personnel.
1) Sum of the four recent changes mentioned in table 9: ranging between -4 centralization on all aspects) to +4 (decentralization on all aspects)
6.5 Patterns of Individual learning and personal skills
While the general formal skill level of organizational staff has hitherto been treated as an intervening (conditioning) variable that may potentially moderate technological impacts on roles and organizational structure, certain more informal skill aspects are also a potential dependent variables that may be influenced by technology directly or by technology-induced organizational change.
Considering the empowerment effects
demonstrated above, we may hypothesize that computers induce higher needs for
personal competencies related to advanced training, self-directed learning,
independent problem-solving and autonomous decision (Baloh/Trkman 2003). As
these informal skills are evidently correlated positively with the level of
formal education, it is paramount to keep these formal skill variables under
6.5.1 Advanced training
First of all, it is evident that in all size categories, extensive computer use goes along with more extensive enrolment in advanced training courses. Once again, almost no additional effects are observed when computers are netted, because the respective correlations- while keeping mostly a positive sign - vanish to insignificance when the PC density is controlled (Table 15). Thus, we may conclude that the major of new skill demands associated with Information technologies have arisen in the early phases where basic knowledge about computer handling and the mastering of software programs had to be transmitted. On the basis of these primary computer skills. employees may then easily learn the technalities of Net related activities like data bank searching and online communication.
While the absolute levels of training involvement are certainly lower when the percentage of unskilled personnel is very high, the correlative relationships between enrollment and computer usage are remarkable similar over all categories of firms. (Table 16). There is only a slight tendency of Network skill effects to decrease (or even turn to the negative) when skill levels are very low.
6.5.2 Self-directed learning
More than any tool from preindustrial ages or any machine originating in classical industrial eras, the computer is a “responsive” technology that engages its user in a looped diadic interaction process that can usually not be observed (end even less controlled) by any outsiders (e. g. collegues or supervisors) (Geser 1989).While such engagements lead to a weakening or dissolution of interindividual interactions, they substitute them by man-machine relationships in which rather intensive and rapid processes of behavioral adaptation occur. In fact, it is hard to use a computer without learning something , because I’m constantly confronted with the causal effects of my immediately proceeding actions. As much of this self-learning occurs unsystematically and unintentionally, it is of course a major basis for formally prescribed processes of self-guided learning: explaining why so many computer skills are acquired by informal “on the job training”, not by any formalized educational procedures. As a consequence, we hypothesize that firms with extensive computer usage give more weight to “self-guided on the job learning” (instead of formalized training) as a major method for upgrading skills.
As seem from Table 17, this
hypothesis holds for only for smaller firms with 100 or less employees, but not
for any larger enterprises. This indicates that larger firms use their resources
for supplementing informal learning with formalized educational courses, so that
more speedy, systematic and homogeneous skill upgrades can be achieved. Contrary to our expectations, Table
18 demonstrates that effects of computer usage on self-guided learning are
highest when the general skill level of employees is rather low (Table 18). From
this, we may draw the double conclusion that computer-induced self-learning also
extends to very low qualificational levels, and that highly qualified workers
are disposed to rely on self-learning anyway, so that computers do not make any
Table 17: Correlations between the importance of "self-guided on-the-job-learning" as a mode of securing individual skills1: and percentage of work places with access to various information technologies: by size of firm.
Table 18: Correlations between the importance of "self-guided on-the-job-learning" as a mode of securing individual skills and percentage of work places with access to various information technologies: by percentage of unskilled personnel.
6.5.3 Autonomous judgment and decision
Given all these effects on autonomous learning and role empowerment, it is to be expected that computer usage goes along with an increased demand for employees with a high capacity to work on their own and to rely on their own judgment and decisions. In fact, the salience ascribed to such skills (by personnel managers) correlates positively with the extensity of computer use, especially in middle-sized firms ranging between 100 and 500 employees (Table 19). In the largest size category, autonomy skills seem to depend critically on the access to transorganizational online communication and the Internet, while the impact of stand-alone computers is much reduced. However, a breakdown by skill categories reveals that most of these correlations may be spurious artifacts originating from the positive intercorrelation between formal skill level and the importance ascribed to autonomy skills. Only in firms with almost no unskilled personnel, a slight significant correlation can still be observed (Table 20).
Table 19: Correlations between the importance of "independent judgment and decision making capacities"1: in the skill profile of average employees and percentage of work places with access to various information technologies: by size of firm.
1 Average value on a scale ranging from 0 (no importance) to 400 (highest importance)
Table 20: Correlations between the importance of "independent judgment and decision making capacities" in the skill profile of average employees and percentage of work places with access to various information technologies:: by percentage of unskilled personnel.
* p < .05 ** p < .01
While acknowledging that the social implications of computers vary according to the ways of their usage, we are nevertheless impressed to what extent the mere presence of IT equipment is associated with many correlates on the level of organizational structures and individual roles.
All these results may be summarized by stating that a broad availability of Personal Computers in a firm goes along with a certain empowerment of the rank-and file. In more specific terms, the following impacts on lower level work incumbents can be observed
1) Job Enrichment. At least in higher skill
environments, employees are typically confronted with a larger variety of tasks:
This job enrichment may well give them the opportunity for acting out a broader
range of interests and skills, for engaging on more learning processes and for
exercising more discretion in the way they divide up their time.
Of course, these correlations cannot provide sufficient evidence for a technological determinism, because at least two alternative interpretations would have to be ruled out:
1) The correlations are artifacts produced by underlying third factors. In particular, it could be argued that the observed empowerment effects are generated by the regularity that firms who have many computers tend to have professional employees with higher skills - employees that would have large work variety and high influence even without any computers. While this contention cannot be fully refuted, it is still made improbable by the fact that empowerment effects are observed in almost all branches and that they do not vanish when the percentage of unskilled personnel is controlled
2) The reverse causality holds: firms with highly empowered employees are in a better position to buy and apply computers. This would imply that empowerment has already been achieved ex ante. However, we have found that highly computerized firms are more likely to have undergone recent changes toward decentralization – what indicates that such changes are in fact a consequence, not an antecedent of technological innovation.
Finally, let’s wonder about the highly consistent, contraintuitive finding that almost all empowerment effects are associated just with the mere availability of personal computers, while the access to computer networks (Intranets, Extranets or the WWW) does not seem to matter. Among many other speculations, it might be argued that in comparison with the PC that has been introduced since the eighties, such computer Networks are too recent to have had hitherto any deep impacts on organizational structures – or that they are even the source of countervailing centralizing impacts which may weaken or neutralize their empowerment effects (e. g. by facilitating more intensive managerial control).
Ang, S. / Cummings, L. L. (1997): "Strategic Response to Institutional Influences on Information Systems Outsourcing," Organization Science (8:3), May-June 1997, pp. 235-256.
Baloh, Peter / Trkman, Peter (2003):
Influence of Internet and Information Technology on Work and Human Resource
Bishop, Libby, and David I. Levine (1999): "Computer-mediated communication as employee voice: a case study." Industrial and Labor Relations Review, 52(2), p. 213-233.
Bloomfield, B. P. / Daniel, A. (1995): "The Role of
Management Consultants in the
Braverman, H. (1974): Labor and monopoly capital: The degradation of work in the Twentieth Century, New York, Montly Review Press.
Bullen, Christine V. / Bennett, John (1990): Learning from User Experience with Groupware. In: Halasz, Frank (ed.): Proceedings of the 1990 ACM conference on Computer-supported cooperative work. October 07 - 10, 1990, Los Angeles, California, United States. p. 291-302.
Carter, N. M. (1984): "Computerization as a Predominate Technology, its Influence on the Structure of Newspaper Organizations." Academy of Management Journal (27:2), April, pp. 247-270.
Dawson, P. / McLoughlin, I. (1986): "Computer Technology and the Redefinition of Supervision: A Study of the Effects of Computerization on Railway Freight Supervisors," Journal of Management Studies (23:1), January 1986, pp. 116-132.
Dean Jr., J. W. / Yoon, S. J. / Susman, G. I. (1992): "Advanced Manufacturing Technology and Organization Structure: Empowerment or Subordination"? Organization Science (3:2), May 1992, pp. 203-229.
Finholt, Tom / Sproull, Lee / Kiesler, Sara (1990): "Communication and Performance in Ad Hoc Task Groups." In J. Galegher, R. Kraut, and C. Egido, (eds.): Intellectual Teamwork. Hillsdale, N.J.: L. Erlbaum, pp. 291-325.
Franz, C. R. / Robey, D. (1986):"Organizational Context, User Involvement, and the Usefulness of Information Systems." Decision Sciences (17:3), Summer, pp. 329-356.
Frenkel, S.J., Korczynski, M., Shire, K.A. and Tam, A. (1999): On the Front Line: Organization of Work in the Information Economy. (Ithaca, N.Y.: Cornell University Press).
Geser, Hans (2002): Toward a (Meta-)Sociology of the digital Sphere. Online Publication, Zürich, http://socio.ch/intcom/t_hgeser13.htm
Geser, Hans (2001): On the Functions and Consequences of the Internet for Social Movements and Voluntary Associations. Online Publication, Zurich. http://socio.ch/movpar/t_hgeser3.htm
Geser, Hans (1989): Der PC als Interaktionspartner; Schweizerische Zeitschrift für Soziologie, Vol. 18, 1989, S.230-243.
Geser, Hans / Bürgisser, Margret (1998): Bald überflüssig - oder befreit für Neues? Sekretariatsarbeit im Wandel (In: Unimagazin, Heft 2, S. 40-44).
Herring, Susan (1992): "Gender and Participation in Computer-Mediated Linguistic Discourse." Paper presented at the Annual Meeting of the Linguistic Society of America (Philadelphia), Jan 9-12. 1992.
Hitt, L. M. / Brynjolfsson, E. (1997): "Information Technology and Internal Firm Organization: An Exploratory Analysis," Journal of Management Information Systems (14:2), fall 1997, pp. 81-101.
Ichniowski, Casey / Kochan, Thomas / Levine, David I. / Olson, Craig / Strauss, George (1996): "What Works at Work." Industrial Relations, Vol. 35, No. 3 (July), pp. 299-333.
Jasperson Jon (2002): Review: power and information technology research: A metatriangulation review. MIS Quarterly Vol. 26 No. 4, pp. 397-459/December 2002. http://www.misq.org/archivist/bestpaper/Jasperson.pdf
Kling, Rob / Lacono, Suzanne (1989): Desktop Computing and the Organization of Work. In Forester (Ed.): Computers in the Human Context. Cambridge, Massachusetts: MIT Press.
Kling, Rob (1996): Computerization at Work (in:
Computer and Controversy: Value Conflicts and Social Choices. San Diego,
Lawler, Edward E. III / Albers Mohrman, Susan / Ledford, Gerald E. Jr. (1995): Creating High Performance Organizations: Practices and Results of Employee Involvement and Total Quality Management in Fortune 1000 Companies. San Francisco: Jossey-Bass.
Lee, Eric (2000): Online-Rechte für
Online-Beschäftigte in den Mitgliedstaaten der Europäischen Union.
Lee, S. (1991): The Impact of Office Information Systems on Potential Power and Influence," Journal of Management In formation Systems (8:2), Fall 1991, pp. 135-151.
Mansfield, Nancy R. / Gabel, Joan T. A. (2003): The information revolution and its impact on the employment relationship: an analysis of the cyberspace workplace. American Business Law Journal 2: 301, Winter.
Mantovani, Giuseppe (1994): "Is computer-mediated communication intrinsically apt to enhance democracy in organizations?" Human Relations, 47(1), p. 45.
Marchand A. Donald, Kettinger J. William, Rollins D. John (2001): Information orientation: the new business performance metric. Oxford University Press Inc., New York.
Martin, Rod (2002): Trust and Technology – the limits of technological control. Paper delivered at the ANZAM / IFSAM 6th World Congress, Gold Coast City, Queensland Australia, July.
Metcalfe, B. (1995): Metcalfe's Law: A network becomes more valuable as it reaches more users," Infoworld, Oct. 2.
Nandhakumar, J. (1999): Virtual Teams and Lost Proximity: consequences on trust relationships, in Jackson, P. (ed.) Virtual Working: Social and Organisational Dynamics (London: Routledge).
Nelson, Merissa (2000): Can Computer-Mediated
Communication Democratize the Workplace? Special libraries associations (SLA).
Newman, M. / Noble, F. (1990): "User Involvement as an Interaction Process: A Case Study," Information Systems Research (1:1), March 1990, pp. 89-113.
Pinsonneault, A. / Kraemer, K. L. (1997): “Middle Management Downsizing: An Empirical Investigation of the Impact of Information Technology." Management Science (43:5), May, pp. 659-679.
Preciado, Regina Lynn (1998): Mouses to the Grindstone, LYCOSNETWORK WIRED NEWS, Aug. 12, http://www.wired.com/news/culture/0,1284,14371,00.html.
Ross, Andrew (2004): No-Collar. The Humane Workplace and its hidden costs. Temple University Press. 2004
Sabherwal, R. / King, W. R. (1992): "Decision Processes for Developing Strategic Applications of Information Systems: A Contingency Approach," Decision Sciences (23:4), fall 1992, pp. 91 7-943.
Schwartz, K. (2000): Too Much Monitoring. http://www.cio.com/archive/091500_mail_sidebar1.html
Sinton, Peter. (1996): "Businesses Wired on Intranets." San Francisco Chronicle. September 26, p. B1.
Sproull, Lee / Sara Kiesler (1991): Connections: New Ways of Working in the Networked Organization. MIT Press. Cambridge, Mass.
Symons, Frank. (1997): "Virtual department, power, and location in different organizational settings." Economic Geography, 73(4), p. 427.
Tan, B. C. Y. / Watson, R. T. / Wei, K. K. (1995): "National Culture and Group Support Systems: Filtering Communication to Dampen Power Differentials," European Journal of information Systems (4:2), May 1995, pp. 82-92.
Tractinsky, N. / Jarvenpaa, S. L. (1995): "Information Systems Design Decisions in a Global versus Domestic Context," MIS Quarterly (19:4), December 1995, pp. 507-534.
Wresch, William C.
(1996): Disconnected: Have and Have Nots in the Information Age.
Rutgers University Press, New Brunswick, NJ.
Given this pervasiveness and flexibility, it is evident that computer technologies are used for a wide range of purposes far transcending narrow work concerns and purely economic goals. Thus, a 1998 survey has shown that employees spend only 75% of their online hours in work related activities, while the remaining time was dedicated to news and entertainment searches and other private concerns. (Preciado 1998)
For instance: controlling the observation of "office hours" becomes pointless when computers allow continuing working at home or on travel.
In his case study about the ESCO Reisen AG, Christian Tanner found that the top managers had sceptical attitude toward the Intranet because they feared that its internal dynamic would lead to a loss of centralized control (Tanner 2000).
/www.edsb.ch/d/themen/internet/index.htm  Metcalfe’s law states that the value of any network equals the square of the absolute number of participants, because every new user also increases its utility for the already existing users (Metcalfe 1995).
Last update: 01. Feb 15