Structure and Strategies of Control : a Replication of Aston Study
In his critical study of bureaucratization Weber listed growth in size and density of population being administered, the growth in complexity of administrative tasks being carried and the existence of monetary economy requiring a more efficient administrative system as the preconditions for the emergence of bureaucracy. Even though Weber’s writings have been widely discredited the bureaucratic form lives.Weber believed that a formal hierarchical structure where each level controls the level below and is controlled by the level above is the basis of centralized decision making. Weber supported controlling by rules as this resulted in decisions made at higher levels to be executed consistently by lower levels.
He was also in support of work being done by specialists based on the skills they possess and that the purpose of the organization was to either serve the agency which empowered it or those within it. Equal treatment of employees was another agenda in his principles.He has concluded that when the organization is in sync with all his principles it will be negatively related to the centralization of decision making. Many writers have put forward alternative views but they too have not denied the validity and utility of bureaucratic model. The only theory which questions Weber’s work is the work of Pugh and his colleagues at the University of Aston.
The well-known empirical study of Aston group has suggested that this one dimensional conception of bureaucracy is no longer useful.Based on a factor analysis of data gathered in 46 work organizations in English Midlands, the Aston researchers concluded that organization structure could be best described by four independent underlying dimensions-structuring of activities, concentration of authority, line control of work flow and relative size of supportive component. Based on these dimensions an organization’s degree of bureaucratization could be found out. They believed that an organization can be bureaucratic in any number of ways with classification of several broad type of organization structures.The study was conducted on various heterogeneous organizations many of which were nonprofit and data were collected by a detailed interview of the head of the organization and the departmental managers. Further investigation was done to ensure that Aston study was suitable for all possible types of organizations.
Thus the new study was formulated keeping the methods of data collection in sync with the previous form. The Aston study thus suggests that formalization and specialization vary together but are in turn independent of decentralization.This was in complete contrast with Weber’s theory. John Child in his replication study collects data samples from 82 British organizations. It is in one way similar to Aston’s study in choosing working organizations however differs in 4 distinct aspects. They being organizations are distributed all over Britain and Scotland (hence they are collectively termed as national samples), samples did not consider subunits or departments hence organizations were considered as one big unit, only business units were considered and last being samples were selected from just six industries.
Industries considered were Electronics, pharmaceuticals, chocolates and sweets, advertising, insurance and daily newspaper. Child categorized these industries based on functions as manufacturing or service industry and based on environment as high variability and low variability. While classifying organizations under 2nd category various statistical data and personal interviews were made use of. All organizations could easily be identified under either of the subcategory of each main category. However daily newspaper industry can be classified under any of the above stated category.One of the important parameter considered was organization size based on number of employees.
The objective of this parameter was to obtain contrast in size within samples, however retaining consistency of size across different industries. Size of each organization was not disclosed in order to avoid identification. It is important to note that mean and standard deviation of size of sample organizations considered by John is half that of Astons’. Also national samples considered by John were homogeneous since none were government organization and all were business units.The most important factor which differentiates John’s study from Aston’s is that complete organizations were considered and not departments or branches.
This choice of not using branches or department and instead choosing entire organization is dictated by a desire to obtain organizational performance data rather than consistency of organizational status. This accounts for differences in findings of Aston’s and John’s research. Between 1960 to 1970 a major attention under the field of organizational theory was given to understanding the organizational structure.This attention to structure can be owed to Webers theory which stated that formalization and centralization are negatively correlated, ie as formalization increases the organization gets more decentralized. This theory raised many questions, beginning from the basic question of if dimensions of an organization might interact with one another.
In 1963 an attempt to answer this question came up from the Aston Study undertaken by Pugh et al which stated 6 dimensions of organization structure namely specialization, standardization, formalization, centralization, configuration and flexibility.This study was just the beginning, and gave a platform for further research on this topic. Pugh et al assigned specific empirical values to each dimension in order to have a standard scale to compare these dimensions across various organizations. Pugh et al used data from 46 different organizations from England to undertake the comparative analysis. The data collected with respect to this specific sample was concentrated on 4 dimensions – structuring of activities, concentration of authority, line control of workflow and relative size of each component.
He then used the same data set to define organizational size, technology and dependence.Pugh et al found organizational size, organizational dependence and network between formalization-technology-location were main predictors of variance in the structural dimension. These results were shared with colleagues via extensive documentation; this was done to gauge the applicability of these results on a much larger sample. Child found an anomaly in Pugh et al’s research which was detected when he tried to reproduce the relationships between dimensions of structure that Pug et al had found. Where Pug et al said that the dimension of structuring and oncentration of authority were independent of one another Child believed that they had a negative correlation. Child arrived at this anomaly by conducting the experiment on a sample which is different from that of Pug et al.
Moreover as there is an increase in formalization with structuring of activities, with defined rules and regulations including detailed job description, impersonal rules etc made the managers more complacent to risk, this is because with formalization, work progresses in a regimented manner with very few mishaps occur, thus the concentration of authority reduced and delegation of work increased.Child proposed that the negative relationship between structuring of activities and concentration of authority was consistent with the original Weberian description of bureaucracy which said that specialization of administrators leads to the delegation of authority at the lower levels. The anomaly reported by Child set off a a series of activities to understand which theory was correct, Pugh et al’s theory of independence formalization and centralization of authority or its contrary proposed by Child. Thus an effort was launched to identify if the findings by Child can be generalized or the results were very specific to his sample only.Mansfield started it of by classifying Child’s sample into six different bands based on size to understand if size played a role in defining the relationship between centralization of authority and the variables underlying the structuring of activities. Mansfield concluded that the size of the sample had no affect and Child’s claim of negative correlation between centralization and structuring of activities.
Next Donaldson et al collaborated on a three-part research note, returning to Child’s suggestion that differences in sample composition between his data set and that of the Aston Group might account for the difference in findings.Classifications of the Aston Group data into various subsets designed to reproduce the distinctions between the Aston and National samples yielded the overall conclusion that the original Aston results (independence of concentration of authority and structuring of activities still held. Thus the anomalous difference between the Pugh et al and Child findings was not resolved. Subsequently, Greenwood and Hinings addressed the inconsistency between Pugh et al. and Child by presenting new data from a study of departments of English local authorities.These data revealed no significant relationships between subscales of standardization and subscales of centralization.
These findings were not supportive of Child’s hypothesis that structuring of authority would provide a control mechanism that permitted decentralization of decision making to lower levels of an organization’s hierarchy. Greenwood and Hinings did not present their results as a conclusive refutation of the Child hypothesis, however. Based on these theories and analysis we can observe that modern organizations have evolved.Earlier in the manufacturing industries where the specialization played a major role, the specialization contributed to a high degree of decentralization, but as the manufacturing industries has evolved with the ever increasing up gradation of technology the power or authority has now been shifted from the specialized administrators with the increase in high degree of formalization as the processes have become standardized. On the other hand the services industry has more degree of decentralization as the services are variable, needs to be customized, more organic in nature and the formalization is also on the higher side.