Enterprise autonomy · Myths and realities Krishna Kumar
Decision-making autonomy may be an
terminant of a business organization's performance. Yet, discussion on the subject has been very limited. Besides, the factors underlying variation in autonomy from one organization to another have not been identified. Lack of empirical studies has led to several stereotypes and myths about autonomy of business enterprises in various economic sectors. Based on questionnaire data and information from interviewing top level executives of enterprises in the public, the multinational and the private sectors in India, the paper identifies sectoral differences in autonomy, the
multidimensionality of the
concept of autonomy, and the situational determinants of autonomy.
Krishna Kumar is a senior core faculty member at the Bank of Baroda Staff College, Ahmedabad.
Vikalpa, Vol. 6, Nos. 3 & 4, July & October 1981
Organizational autonomy may be an important determinant of management effectiveness and organizational performance. The fortunes of a company may swing with the business cycle if its management is not permitted to diversify into contra-cyclical products. Similarly, for a company which is already a low performer, low autonomy in matters related to motivational aspects such as compensation and promotion of executives becomes critical for the survival of the company. On the other hand, higher autonomy if used judiciously, may lead to higher and steady prosperity of the company. Ironically, while divisional autonomy has drawn the attention of control system designers, autonomy is often overlooked in comparing the performance of two or more business organizations. Companies are formal organizations, and legally they are autonomous. But this legal autonomy is often misinterpreted as behavioural autonomy vis-a-vis the owners of the controlling interests. It is necessary that one has a clear idea of the real, behavioural autonomy available to the management of such legally autonomous organizations, before one can understand and analyse the management and functioning of such organizations. Management literature on issues of organizational autonomy is scanty. The discussion has, by and large, not gone beyond relating autonomy to control and public accountability. Dutt (2), for example, has analysed the reasons underlying the need for autonomy for better management, highlighting its limitations and suggesting that autonomy be coupled with purposive control. The twin issues of autonomy and control have also been discussed by lshwar Dayal (1) who suggests that controls should be designed with respect to task
and environment, as unnecessary controls lead to defensive behaviour. Kamath (3) while em~ phasizing the need for maximum autonomy, has highlighted the underlying conflict between autonomy and public accountability and advises a right balan·ce between autonomy and control. Occasionally discussion has also touched upon the effect of autonomy on performance (5) but it has never gone beyond the speculation that reduced autonomy has an adverse effect on performance. Some authors have attempted to explain the reasons underlying variation in organizational autonomy. Manendra Mohan (6), for instance, has attributed it to the style of functioning and thinking of the top decision makers, and tolerance for decisions taken by others, especially subordinates. However, there may be additional situational determinants of autonomy. Further, none of these works seems to be based on empirical research, making it difficult to assess the extent to which the arguments are generalizable. Besides, their prescriptive undertones are apparent. The only exception to this seems to be the study conducted by Pugh (7). The scope of the study, however, was restricted to the relationship between organizational autonomy and the internal structure of the organization~ The study thus has limited utility for discussion of the issues at the corporate level, or for identifying the situational determinants of organizational autonomy. Certain types of ownership may provide more autonomy to the management of the organization than others. In the context of autonomy, organizations can be considered as of two types: one, those that can be called integrated systems, and the other, those that are differentiated systems. In the case of the former, the owners or the controlling interests themselves constitute the top managememt, and their decisions on all matters are final. A proprietorship can be considered as an integrated system. In differentiated systems, the management and the owners or the controlling interests are different sets of persons, the organization being run through representative management. The concept of autonomy
is relevant to these organizations. Public sec tor companies are examples of differentiated systems. But in the private sector also, subsidiaries of multinational companies in India, having parent companies outside the country, could also be considered as differentiated. The situation may not be very clear in the case of domestic private sector companies. While most domestic private sector companies are integrated system organizations, the exceptions may be those domestic private sector companies that are components of business houses. In such cases the companies in a group may either be differentiated or integrated systems depending on the policy of the business house in respect of the autonomy granted to the enterprises in the group. Discussion on autonomy in the context of Indian companies has remained by and large confined to public sector companies. Such exclusive discussions have limited use, for it is difficult to say how much autonomy should be given and how much of it exists at present. Further, one also does not know whether these companies are in any disadvantageous position vis-a-vis those with whom they are compared for performance. The underlying assumption of total autonomy in the case of private sector. companies is questionable when they are viewed from the differentiated and integrated systems points of view rather than from the viewpoint of their legal independence. Little information is available on the behavioural autonomy of private or public sector units to permit any sweeping generalizations about which sector has more autonomy. Another closely related but neglected aspect of managerial autonomy is the concept of autonomy itself. So far, autonomy seems to have been considered as an unitary concept. But in fact different organizations may be autonomous with respect to different matters, and it may be more fruitful to compare, say, public and private sector enterprises not just on total autonomy but also on autonomy regarding various facets of the management of the enterprises.
$cope of the study This paper describes a research study which attempts to provide answers to questions such as: 1. Does the legal autonomy of companies also imply behavioural autonomy? 2. Is the question of autonomy relevant only to the public sector? If not, where do companies in the public sector stand vis-a-vis those in the private sector? 3. Should autonomy be understood as an unitary concept or as a multidimensional variable? 4. What are the determinants of the level of autonomy? The study is in two parts. The first is descriptive, depicting the level of autonomy enjoyed by the managements of companies in different sectors in India. This part largely responds to the first three questions. The second part expounds the factors that are responsible for low or high level of autonomy. The discussion in the first part is based on the data gathered through questionnaires administered to the chief executives (or equivalent) of the companies studied. The interpretation is based on the analysis of the data. The discussion in the second part is based on the information gathered through interviewing the chief executive and also some other executives of the companies studied, as well as officers of the Bureau of Public Enterprises and the Ministries to which the public enterprises included in the study are attached. A total of 30 companies were included in the study. However, while interview information could be gathered from almost all, the questionnaire responses could be received only from 23 companies. Seven of these were in the public sector, 7 others were Indian subsidiaries of multinational companies, and the rest were in the domestic private sector. The companies operate in sophisticated technology industries, namely chemicals and fertilizers, electrical equipment and cables, and engineering (as per Bombay Stock Exchange classification). The sales turnover of the com-
Vol. 6, Nos. 3&4, July& October 1981
panles studied ranged from Rs. 10 crore to about Rs. 1 00 crore in 1976. None of these companies was less than 10 years old. Questionnaire details Seven questions were asked from the company executives to get the information on the autonomy enjoyed by the management of the company vis-a-vis the owners. Four of them pertained to the marketing function (decisions regarding choice of clientele, product, distribution channels, and promotion); two pertained to the finance function (capital investments and operating budgets); and one to the personnel function (recruitment and compensation of senior executives). The information was sought through 7 point scales (for details, see 4). Data analysis The data analysis was done at two levels. At the first level correlations between 7 items of autonomy were calculated, which indicated the degree to which the level of autonomy in each area is related to the level of autonomy in the remainder. At the second level of analysis companies in each ownership group were compared to those in the others in each of the 7 items of autonomy. Using the actual scores given by each company the Mann-Whitney test was conducted to find out whether the companies belonging to any particular group were (statistically) significantly different from those in the other groups. Tables 1 and 2 show the findings of the study. Findings of the study Table 1 gives the correlations among the 7 components of enterprise autonomy. Statistically significant relationships were discernible between autonomy in clientele selection and in product promotion policy decisions, as well as between autonomy in product selection and autonomy in distribution channel decisions. The relationship between autonomy in clientele choice and in product selection; in product selection and in promotion policy decision; in promotion policy decision and in distribution channel decisions, were surprisingly low.
A notable fact that emerged -was that
autonomy in all the key decision areas of marketing, mainly product choice, promotion, choice of distribution channels, and client selection does not necessarily co-vary. The other significant relationships observed were between autonomy in setting operating budgets and in clientele selection as well as in promotion policy decisions. Autonomy in capital investment decisions on the other hand was (statistically) significantly related to autonomy in clientele selection and in setting operating budgets. While not all the significant relationships 9bserved were easily explainable, the undisputable fact that emerges from the analysis is that the level of autonomy in one function or one policy area need not necessarily match the level of autonomy in the other functions or policy areas; that is, low autonomy in one management area might simultaneously exist with high autonomy in another function. These findings suggest that for any discussion on the subject of autonomy, one should not consider autonomy as an unitary concept, and should not analyse the issues of autonomy and their implications in terms of total autonomy. Instead, one should be more specific, and should ideally focus on autonomy in each different area of management. Table 2 indicates the levels of organizational autonomy on the 7 areas described earlier across the three sectors. Out of the 7 items, pronounced sectoral differences showed up in autonomy concerning senior executives' recruitment and compensation, where public sector clearly had the least autonomy. In the rest of the 6 items no pronounced sectoral differences are discernible although a few differences between pairs of sectors are seen, as for example the generally greater autonomy in the private sector concerning client selection and choice of distribution channels as compared to the public sector, and the generally greater autonomy in the private sector concerning product selection, fixing of operating budgets, and capital investment decisions as compared to MNCs. The above findings suggest that although
sectoral differences are realities, they are by no means overwhelming. However, the differences between the public sector and the two private sectors on autonomy in the selection and compensation of senior executives may be critical, for it may lead to less flexibility in motivational strategies in the public sector, which may affect negatively enterprise performance as well as managerial job satisfaction. The findings in Table 2 also belie the popular myths prevalent about the organizational autonomy of the private sector companies (the domestic private and the multinational subsidiary). These companies have been, by and large, never analysed by the level of autonomy they really enjoy. The actual levels of autonomy enjoyed by them (see Table 2) indicate that in some of the autonomy items such as capital investment decisions, product selection etc., many of the multinational subsidiary companies may be having as low autonomy as their counterparts in the public sector. The same holds true for some of the companies belonging to the business houses in the domestic private sector. Indeed, the same is likely to hold true for all organizations that are parts of super-systems. It may, therefore, be more useful to classify domestic private sector companies into two categories, namely independent or autonomous companies (having no super-system), and (those which are part of a large system) non-independent companies. The findings reported in Tables 1 and 2 thus provide the following information on the organizational autonomy of the domestic private, public, and the multinational subsidiary companies sectors in India: i) The legal status of an enterprise is only thinly related to actual operating autonomy. Although private sector enterprises on an average may have more autonomy than public sector enterprises, there may be many public sector enterprises that may be enjoying as much or more autonomy than a significant number of private sector or multinational enterprises. The sectoral differences are by no means overwhelming.
Organizational autonomy may vary considerably across areas of management. Low autonomy in a particular area may not necessarily mean low autonomy in other areas. For making valid public-private sector comparisons, companies in the domestic private sector could more usefully be classified into two broad categories, namelY, those which are not part of a business group (i.e., no super-system) and hence the organizational autonomy factor is irrelevant in their context; and those which are non-independent (part of business group) and hence may have low autonomy in one or more functions, areas or issues. Within each sector, companies exhibit a cE>nsiderable range of autonomy in most management areas. Thus, low or high autonomy cannot be easily attributed to any particular sector. There may be much more significant situational determinants of the extent of autonomy than merely who owns an enterprise.
Situational determinants of enterprise autonomy The question of low autonomy is relevant only for those organizations which are part of a super-system. A super-system can be defined as an organization or a set of organizations that has or have, direct or indirect control over the management of a particular organization. For example, in the case of the public sector, the government, particularly the controlling ministry, forms the super-system; for a multinational subsidiary, the parent multinational constitutes the super-system; and for domestic private sector companies the business house to which they belong represents the super-system. The level of organizational autonomy enjoyed by an organization may depend on certain specific characteristics of the supersystem and the relationship between it and the organization embedded in the system. From Table 2 it is apparent that different companies in a sector may enjoy rather diffe-
Vol. 6, Nos. 3 & 4, July & October 1981
rent levels of organizational autonomy. This indicates determinants of an organization's managerial autonomy additional to the type of ownership. The interviews conducted with top management executives (usually the chief executive) of the companies inciuded in the study indicated several situational determinants of enterprise autonomy. The interviews suggested that the factors determining the level of enterprise autonomy were: the rationalization or standardization of policies by the super-system's management for all the constituent organizations; the integrated manner of resource allocation in the supersystem for achieving its objectives; the dependence and importance of the organization on its system; and the importance of a particular policy to its super-system.
The rationalization of policies by the supersystem Rationalization in a P.Oiicy imposed
by the super-system on its constituent organizations is intended to achieve parity or uniformity among all of them. Such rationalization implies the reduction of autonomy of the constituent organizations. As was seen, the public sector companies showed a distinctively low autonomy in respect of decisions related to the recruitment and compensation of senior executives. Such low autonomy seems to be the outcome of the attempt on the part of governmental super-system to rationalize the policies concerning the senior executives' pay and other benefits in public sector undertakings. The concern of the government seems to be that there should not be unwarranted disparities among public sector organizations. An important factor influencing the degree of rationalization attempted by the supersystem may lie in its approach towards the management of the constituent organizations. Does it want to run all of them in a closely knit manner or through a loose federal arrangement? In this study sharp contrasts were noted in the case of organizations belonging to large Indian business groups. One of these was very closely managed, to the- extent that it would be difficult to understand and analyse a company in the group without knowledge
of the policies and practices of the other companies of the group. Very close liaison was maintained between them. Many of the decisions were taken jointly. Some companies seemed to find it more convenient (for legal or financial reasons). to float another company than add another division to it, thus leading the first step towards the growth of a supersystem. When a super-system originates from such considerations then each company in the system may be very closely managed by the latter, resulting in low autonomy of the constituents. In complete contrast to this, there seemed to be cases in which the super-system was only of face value. No real top management of the system seemed to exist. Such a situation could be seen in some of the large business houses in the domestic private sector that are in the process of disintegration, the empire having been divided among the family members. The mini-empires might eventually control the companies belonging to them. In such situations companies clubbed as a business house may indeed be behaviourally quite autonomous, despite the legal presumption of not being so. There could also be cases where such cracking up has not taken place but the sheer number of companies, their diversity, or size may be so great that the super-system does not, or cannot, control them tightly. It may be worthwhile to identify the limitations of the degree to which rationalization can be attempted by the super-system. Consider the rationalization of the wage structure attempted by the super-system in the case of public sector companies. Could the same attempt be made by the multinational supersystem (parent company) of an Indian subsidiary that has several such subsidiaries in different parts of the world, operating in rather different labour markets? Surely, it would be difficult. Similarly, attempts at rationalization of financial policies of the subsidiaries by the parent may be restricted because of diverse tax regulations, financial markets, etc., of the countries in which the subsidiaries are operatl[ing. The heterogeneity of the input and output
markets of the organizations forming part of a super-system seems to limit the extent to which rationalization may be attempted by the super-system. The integrated manner of resource allocation by the super-system At times, the top management of the super-system may have an optimization orientation. With limited resource, it tries to achieve its objectives to the maximum. This is done through an integrated or centralized resource allocation to various constituent organizations. The super-system, while deciding the resource allocation, considers the total resources available to it, the requirements of the sub-systems, and the optimal achievement of the super-system's objectives. For example, suppose the supersystem's basic objective is to increase profitability then, the capital investment decisions will favour those companies of the supersystem whose returns on investment are comparatively higher. In the process the other companies may not be permitted further capital investment even if their managements have generated profits or have enough internally generated resources to meet their investment requirements. Thus, an optimization orientation in the super-system's management may restrict the autonomy of the constituent organizations with respect to capital investments. Criticality of a policy for the super-system Sometimes a particular policy assumes critical importance for the super-system and if freedom is given to the constituent organizations in respect of certain functions or activities, it may jeopardize the effectiveness of the policy. For example, as indicated by managers in several multinational subsidiary companies, borrowing by the subsidiaries is restricted or a particular debt/equity ratio, even if comparable domestic companies (both in the public and the private sectors) have a much higher ratio, because it would adversely affect the debt/ equity ratio of the parent company, which in turn may affect the parent company's standing in the financial markets. Thus, because of the criticality of the debt/equity ratio for the parent company (super-system) autonomy of the subsidiary company in borrowings is restricted. Vikalpa
Dependence of the organization on the super-system The dependency of the organiza-
tion on the super-system may also cut into the autonomy of the organization. Those public sector companies that have surpluses and reserves enough to meet their capital investment requirements and rely only marginally on funds from the government, seem to be more autonomous than those which have to rely on the government for meeting their capital investment_ requirements. Autonomy tends to be lower in the case of organizations that are comparatively large, as the stakes for the supersystem in their cases are high. The situation worsens for those organizations which are not perceived as good performers by the supersystem. The dependency of the organization further increases if it is assigned the role of a captive unit for other parts of the super-system. It was seen that those public sector undertakings which were supplying products to one or the other government departments or agencies were having considerably lower autonomy as compared to those which were not captive units. Thus, the higher the dependency of the organization on the super-system, the lower tends to be its autonomy.
The importance of the organization to the super-system All the constituent organizations
are not equally important to their super-system. However, those that are more capable of achieving the super-system's objectives may receive more attention. Comparatively good performers often tend to be the blue-eyed boys of the super-system, and they tend to get most of their demands met by the parent body. Unlike the factors which influence the level of autonomy in specific issues and policy matters, the importance of the organization to the super-system tends to have a global effect on the autonomy enjoyed by the organization. The effect does not seem to be always unidirectional. The importance of the organization may lead sometimes to higher or to lower levels of autonomy, depending upon whether autonomy facilitates or inhibits the level of achievement of the super-system's objectives.
Vol. 6, Nos. 3 & 4, July &·October 1981
Implications of the findings In the first part of the study, major sectoral differences in respect of the organization's autonomy were sought. In the second part of the study, based on interviews with managers and officials, several factors that influence the level of autonomy enjoyed by an organization have been identified. It is obvious that the difference between legal autonomy and behavioural autonomy is a real one. Companies set up as independent organizations as per the law may not be behaviourally autonomous. This fact suggests that organizational performance may not be a good measure of management performance without considering the degree of autonomy provided to an organization by its super-system. Thus a comparison of the economic performance of any two organizations from the same or different sectors for evaluating managerial performance may be inappropriate, as also suggestions to improve management by suitably changing the top management of the organization, in the absence of comparable autonomy of the two organizations. Indeed, even the super-system of the organizations should not use these measures to evaluate managerial performance of the constituent organizations, if they grant unequal autonomy to them. It was noted that public sector companies generally have lower autonomy as compared to multinational subsidiaries and other domestic private sector companies. The differences, however, were not very large, with the exception of autonomy with respect to the senior executive's recruitment and compensation. Indeed, in most of the studied areas of autonomy, all the three sectors had companies having high as well as low autonomy. Thus, public sector companies cannot take the shelter of low autonomy as the sole defence for their poor performance compared to those in the other sectors. The findings also expose the myth that all private sector companies are behaviourally autonomous, by revealing that in many areas, multinational subsidiaries and domestic private sector companies enjoy low autonomy~like public.:sector companies.
The study has also highlighted that the level of autonomy in one area does not necessarily imply low autonomy in other areas. Low autonomy in one may, at times, exist with high autonomy in other areas. It is thus necessary that while analysing the level of autonomy in reference to a particular organization, one bears in mind the fact that autonomy is a multi-dimensional phenomenon. The question of autonomy should be thus considered in terms of how much autonomy and in which specific areas of decision-making. Extending the argument a little further, the issue of autonomy may be discussed with reference to its criticality. Autonomy in every area may not be equally critical. Its criticality is likely to depend upon the internal and external environments of a particular organization or organizations. For example, in an organization with morale problems, low autonomy with respect to motivational practices may be critical for performance; in a company with cyclical product/product lines, low autonomy with respect to product selection may be critical for performance. While deciding the level of · autonomy to be allowed, the specific characteristics of the concerned organization assume great importance.
The fact of large intra-sector variations in autonomy suggests that on the question of autonomy one should not asssume that all the companies in a particular sector are alike. Before framing public policies, one needs to enquire about the actual level of autonomy enjoyed by the enterprises rather than simply make a priori assumptions. The study also highlights the several situational determinants of the level of autonomy of an embedded organization. These may be as, or more important than the personalities of managers of the super-system and the embedded organizations. These should provide valuable help in understanding why intra-sectoral differences in autonomy exist, and in tentatively predicting how much autonomy a particular embedded enterprise will have. A better understanding of these situational determinants of autonomy can help a management secure greater autonomy by appropriately manipulating the situations that affect autonomy. For· example, helping the super-system achieve its objectives may well earn for the organization greater autonomy than before, as also less dependence on the super-system for resources or better compliance with policies critically important to the super-system.
Table 1 Correlation coefficients of management's autonomy from owner in different functional areas
1. Clientele selection 2. Product selection 3. Promotion policy 4. Distribution channel 5. Senior executive's compensation. promotion, etc. 6. Operating budgets 7. Capita I investment
0.03 0.76* 0.26 0.39 0.49* 0.61 *
O.D7 0.12 0.01
0.23 0.24 0.47* 0.44
0.19 0.16 0.30
*represents coefficients significant at 0.05 level.
Table 2 Autonomy analysed by extent, sector. and areas of management
1. Autonomy in clientele selection decisions
Autonomy in product selection decisions
3. Autonomy in promotional policy decisions
4. Autonomy in distribution channel decisions
5. Autonomy in senior executives' recruitment. compensation, and promotion 6. Autonomy in operating budget decisions
7. Autonomy in capital investment decisions
High Medium Low High Medium Low High Medium Low High Medium Low High Medium Low High Medium Low High Medium Low
7 2 0 4 2 3 8 0 1 5 3 1
3 1 3 2 2 3 5 2 0 1 3 3 0 1 6 3 0 4 1 3 3
4 3 0 0 4 3 4 3 0 4 1 2 6 1 0 0 5 2 0 5 2
4 2 3 6 2 1 4 3 2
Pvt.-indicates domestic private sector Public-indicates public sector MNC-indicates Indian subsidiaries of multinational companies.
Vol. 6, Nos. 3 & 4, July & October, 1981
sector. and multinational subsidiary companies in India,"
in public sector," Lok
doctoral dissertation, liM. 1980. 5. Mohan. Manendra. "Public sector: Autonomy of in-
1. Dayal. lshwar. "Autonomy and control sector;· Lok Udyog, Vol. 3. No. 3, 1969. 2. Dutt. R. C.. "Autonomy Udyog. Vol. 1. No. 8. 1967.
dividual enterprises," Economic Times, March 31. 1981.
3. Kamath, N.J .• "Accountability and operational autonomy in public sector," 1968. 4. Krishna
Lok Udyog. Vol. 2. No. 9.
"A comparative study of general between
6. - - - . "Public sector: Executives and supervision," Economic Times, April 1. 1981.
7. Pugh, D. S. eta! .. "The context of organization structure," Administrative Science Quarterly, Vol. 14, No. 1, 1969.