Burt 1980

Autonomy in a Social Topology

Abstract

To define and illustrate a concept of "structural autonomy"

the pattern of relations defining a network position

(Oligopoly, Group-affiliation)

-> 8 Hypotheses

1) Relations -> Places in a social structure where cooptive relations should appear as well as where they should not  -> Autonomy

2) Effective cooptation -> The increase in autonomy

1) Firms in manufacturing industries of the 1967 American economy as structurally equivalent actors # SE? really? 

2) Total profits in an industry as a result of the relative autonomy of firms in separate industries # Proper? Power

3) Autonomy -> Relative industry profits and strategies for coopting other firms # Inter or Intra industry

Result: Industries with high structural autonomy tend to have high profits # Autonomy is the feature of what? a structure or a firm?

Merger: coopt constraints on the industry's structured autonomy # Therefore, autonomy is the space of a system

Relations -> Division of labor (Independency) -> Roles  & Jointly occupied statuses (<- Interlocking role-sets)

Social structure: Static patterns of relations between statuses in the system

How does occupying a particular status in society determine freedom from constraint?

How does the pattern of relations defining an actor's "position" in a system determine his autonomy, that is, his ability to pursue and realize interests without constraint from other actors in the system?

* Autonomy as a structural concept

- The social topology of relational structure

The social structure of a system of actors as a "social topology." 

Position, Distance, and Equivalence (see Burt 1976, 1977b, 1977d, 1978, for details).

A "position": a pattern of relations to and from an actor within a system of actors

A "distance": similarity of relations with every actor in their system

"Structurally equivalent": zero or negligible distance, jointly occupying a single position as a system status/role-set.

N actors / M statuses

"Blockmodel" of the system's social structure (with no residual group)

The position J: Aggregation of row J and column J (focal block's relation with all the other groups)

What is it about the pattern of relations defining position J in figure 1 that could yield its occupants high autonomy relative to occupants of other positions in the system?

1) with the autonomy of occupying a position, not with the autonomy of particular actors per se, 

2) with the relative autonomy of occupying structurally nonequivalent statuses, not with an absolute level of autonomy per se.

- Aspects of autonomy

Two basic aspects to the idea of being autonomous within a system of actors, either of which can be treated as embodying the overall concept of autonomy.

1) Autonomy via oligopoly

Political economy: Collusion. {Adam Smith's discussion in The Wealth of Nations ([1776] 1937), }

# yj1: Centralization within a block J 0 < y < 1 (oligopoly)

The autonomy of actors in an oligopoly, with regard to the constraint of supply and demand, is illustrated by their ability to raise the market price for their "commodity" far above the natural price (cf. Stigler 1964; Shepherd 1970, pp. 11-47).4

An oligopoly to eliminate competition within their position and, accordingly, to escape the constraints of supply and demand.

# Competition within the industry

Measure of oligopoly

The relations among actors jointly occupying a status in a system: Element (J, J)

The actors jointly occupying position J will be able to escape the constraints of supply and demand imposed by actors in other positions and, accordingly, will be "autonomous" within their system, to the extent that among persons, or corporate actors, occupying the position there exists an oligopoly (few competitive decision makers) or, in the extreme of centralization, a monopoly (a single decision mak

0 < y < 1 (oligopoly)

Centralization: The ratio of total sales by some number of the largest firms in a sector / the total sales by all firms in the sector

(In sectors of an economy)

+ Freeman (1977) suggests measures that capture the extent to which communication among a set of persons must pass through a single "central" person. # Collusion is not the oligopoly?

2) Autonomy via group-affiliation

Offset

Collusion: Undifferentiated, cohesive system

-> Actors in each sector are constrained by their lack of differentiated relations to other actors in the system.

Group-affiliation hypothesis

For both Durkheim and Simmel, differential freedom from constraint by society occurs as a result of differential complexity in an actor's relations to other actors.

Durkheim: Balancing of forces between occupational groups as statuses created by a division of labor and by moral authority of government.

Simmel: the competition among groups linked to an actor for his attention and conformity

-> Autonomy is high for actors occupying a position with many conflicting group-affiliations and low for those occupying a position affiliated with only one other position 

Actors jointly occupying a position that forms an oligopoly are subject to the constraints of the oligopoly. These constraints, in contrast to autonomy via oligopoly relative to other positions, serve to limit the autonomy of the individual wishing to deviate from other actors occupying his position.

The constraint on individual actors of being too strongly integrated into a group of similar others

The relative autonomy of separate position

"Group-affiliations" : Affiliations to actors in other positions rather than to structurally equivalent actors

The manner in which actors jointly occupying a status are related to actors occupying other statuses

Actors jointly occupying position J will be able to balance demands from other actors and, accordingly, will be "autono- mous" within their system, to the extent that the pattern of relations de- fining position J ensures high competition among those actors who interact with the occupants of position J.

Autonomy via group-affiliation emphasizes two characteristics of the pattern of relations defining a position. 

2-1) Actors occupying position J will have high autonomy to the extent that they have relations to many other statuses rather than with only one other: the extent to which actors occupying a status have diversified relations with other statuses

# More relationship with others 

2-2) Actors occupying position J will have high autonomy to the extent that the statuses with whose occupants they do have relations are not oligopolies: the extent to which they have relations only with statuses that are too poorly organized to make collective demands

# When others are not centralized

# yj2 = sum of xji (when i is not j): complex relationship between j and i

= each i's centralization * (relative relation strength of i for j^2  + relative relation strength of j for i^2)

(Undifferentiated) 0 < y < 2 

2: when actors occupying position J initiate interaction only with actors occupying a single other position and are themselves the object of interaction from actors occupying a single other position and these position(s) are perfectly centralized (Y1 = 1).

0: the occupants of position J have relations only with actors jointly occupying very decentralized positions (y = 0) and have no relations with actors whose positions are perfectly centralized (y= 1).

- Structural Autonomy

Network concept of autonomy: oligopoly focusing on the relations among structurally equivalent actors occupying a status + group-affiliation focusing on relations linking those actors to other statuses.

High structural autonomy

1) There is high centralization among occupants of the position such that they form an oligopoly

2) Their relations to other statuses are diversified and exist only with statuses that do not themselves form oligopolies

3) The first two conditions occur simultaneously as an interaction effect

Autonomy ~ + Oligopoly - Undifferentiated  (Lack of conflict group affiliation) + Oligopoly * Undifferentiated

: contains information on where cooptive relations should occur

Cooptation: the process of absorbing new elements into the leadership or policy-determining structure of an organization as a means of averting threats to its stability or existence

Hypotheses 1-3: The likelihood of actors establishing formal authority 

The relative autonomy of occupants of separate positions in a system

H1: Oligopoly should lead to autonomy

H2: Conflicting group-affiliations should increase autonomy

H3: Simultaneously forming an oligopoly and having conflicting group-affiliations leads to autonomy

Actors jointly occupying position J will have high structural autonomy to the extent that their relations ensure low competition among occupants of position J and high competition among the nonoccupant actors who interact with the occupants of position J.

"Cooptive" relation

Wji: an informal relation giving the actor(s) occupying position J some effect on the decisions made by actor(s) occupying position I

Informal relation: a relation that is relatively dependent on, or at the discretion of, the individuals performing it. 

Formal relation: a role or technical requirement that is imposed on the individuals performing the relation

A cooptive relation should appear whenever such a relation increase the autonomy of occupants of position J

An increase in monopoly will lead to an increase in autonomy, as long as the occupants of position J are not strongly constrained by actors in other position 

There should be a significant level of cooptation among the occupants of a network position as long as they are not more constrained by actors in the system than the average occupants of positions in the system

# The condition when oligopoly leads to the autonomy. Therefore, within cluster relationship increase the autonomy in this time

Hypotheses 4-7: The likelihood of actors establishing informal, potentially cooptive relations with groups of actors threatening their autonomy

Where cooptive relations should occur among the occupants of a network position

H4: Structurally equivalent actors will always have cooptive relations with one another as long as they are not subject to high constraint from actors outside their position. (No constraining from outside leads to cooptive insiders in J)

cooptive relations with actors in other positions

H5: Unless the actors occupying position J are poorly organized such that the above partial derivative is positive, there should be a significant level of cooptation by position J's occupants of the occupants of each other position I which places a significant constraint on the autonomy of position J (Differentiated J leads to cooptation with constrainer I)

H6: If the occupants of position I do not constrain the autonomy of actors occupying position J, there is no need for actors occupying position J to establish cooptive relations with position I's occupants. (No constrain, no cooptive)

H7: If the occupants of position I show high autonomy of actors occupying position J, there is no need for actors occupying position J to establish cooptive relations with position I's occupants. (No constrain, no cooptive)

This variability in successful cooptation means that autonomy estimated from formal relations alone will be erroneous in a predictable manner

H8: the extent to which true autonomy (aj) exceeds the level of autonomy erroneously estimated when cooptive rela- tions were ignored (call it aj) should have a positive slope (,a) when regressed over the extent to which occupants of position J have successfully coopted threats to their autonomy (d[aj]) (Successfull cooptation -> More autonomy)

- Numerical Illustration

* A STRATEGIC OPERATIONALIZATION

Manufacturing industries in the American economy as structurally equivalent firms whose pattern of transactions with other firms acting as suppliers and consumers has inherent in it some level of struc- tural autonomy reflected as the relative level of profits obtained by firms in the industry

1) Perhaps most important, data are readily available on industry profits as a reflection of autonomy. 

2) Data are readily available on the economic transactions defining each manufacturing industry as a network position in the economy. 

3) Economic transactions as formal relations are clearly distinct from a variety of informal-and potentially cooptive- relations involving industries in other sectors of the economy

- Industry Profits as a Reflection of Industry Autonomy

The level of profits obtained by a firm within an economy thus provides a clue to the lack of market constraint confronting that firm. Those firms obtaining the highest profits should be the firms with the highest structural autonomy in the market.

- Manufacturing Industries as Jointly Occupied Network Positions

- Cooptive Relations Involving Firms in Manufacturing Industries

The option of creating and destroying informal social relation

: diversification through merge

* ANALYSIS 

The effects are weaker for the 20-industry level than they are for the 335-industry level; however, all coefficients are in the expected directions, and the results for the 335-industry level strongly support hypotheses 1, 2, and 3.

(20 / 335)

As expected under hypothesis 1, oligopoly has a positive effect on profits.

# Oligopoly -> +**

High values of Y2 are associated with low profits, as expected under hypothesis 2

# Group-affiliation -> -**

# Less relationship with others # When others are centralized

Hypothesis 3 receives the weakest-albeit statistically significant-sup- port. 

3-1) At the 335-industry level of aggregation, concentration and group- affiliation are nearly independent (r = .073). Even so, there is a significant interaction effect from the two variables at less than the .05 level of confidence (unstandardized /3, is .792). 

3-2) At the 20-industry level, in contrast, concentration and group-affiliation are highly correlated (r- .333), and their interaction effect on profits is negligible.

# Interaction effect *.

Hypothesis 4 says that firms in an industry will establish cooptive relations with one another as long as they are not too constrained by firms in other sectors

# Noy too much constrained by outsiders -> Cooptation of insiders*

Hypothesis 5 says that firms in a sector having a negative effect on the structural autonomy of firms in an industry will be the object of cooptive relations, as long as firms in the industry are not too disorganized.

# When not too much oligopoly in J, I's constraint on J -> Cooptive relations with I .

In short, the merger data demonstrate a slight, but hardly overwhelming, asymmetry corroborating the cooptation hypotheses as stated in the text.

# Research on the cooptive uses of corporate boards of directors has ex- tended the domain of potential cooptees to include nonmanufacturing sectors (Burt, Christman, and Kilburn 1979). The results are encouraging. Ownership ties, direct in- terlock ties, indirect interlock ties through financial institutions, and multiplex cooptive ties between corporations all have a statistically significant tendency to occur in the presence of market constraint, as predicted by hypothesis 5.

Hypotheses 6 and 7 say that firms in a sector which does not constrain the structural autonomy of firms in an industry will be ignored in the industry's cooptive strategies.

# No I's constraint on J -> No cooptation *

Hypothesis 8 says that the errors made in predicting profits in table 4 have a specific meaning.

# The interindustry merger relations could be eliminating the bulk of market constraints in manufacturin

# The correlation between expected increase (d[aj]) and the difference between observed and predicted profit margin (yjo - ?jo) is negligible (r - .01)

The lack of support for hypothesis 8 is, to some extent, a result of support for the other hypotheses. 

While significant merger relations do not occur whenever there is a significant market constraint (as evidenced by the statistically negligible support of hypothesis 5), virtually all the largest market constraints confronting each industry are covered by a significant merger relation

# Thus, and in opposition to hypothesis 8, successfully coopting market constraints appears to be an attribute of all industries rather than a variable distinguishing industries in terms of their ability to obtain profits in excess of the profits to be expected from their relative levels of structural autonomy.

# Inhibit Homoscedasticity

* CONCLUSION

How the pattern of relations defining a network position "frees" occupants of the position from constraint by other

The context of a system stratified across structurally nonequivalent statuses/role-set

Scope condition

The most promising application: systems where there is a clear separation of formal from informal, potentially cooptive relations

Example

1) In a corporate bureaucracy

1-1) the relative discretion allowed to executives occupying positions in the corporation

1-2) Informal friendships to develop where constraint on each position is high. 

2) Between corporate bureaucracies

2-1) Structural autonomy predicts the relative freedom of corporations in sectors of the economy to set prices independent of other sectors 

2-2) Diversification, joint ventures, interlocking directorates, etc., to develop where constraint on each sector is high.

Question

How groups should be interconnected and why, in terms of the constraints corporations place on one another as a result of their network of transactions.

The network of transactions -> Constraint -> Space