This paper builds on previous work by Ouchi who developed a set of control archetypes to explain variety in the area of control. Speckle develops a more comprehensive framework in terms of control archetypes that he refers to as a transaction cost theory (TCE) of management control (MC). My summary follows the paper's five sections:
    1. Introduction,
    2. An Introduction to TCE,
    3. TCE and MC: Some Preliminary Considerations,
    4. A Transaction Cost Theory of MC, and
    5. Summary and Conclusion.

1. Introduction

The problem of explaining management control (MC) structure variety has been addressed in contingency theory and agency theory research, but both theories have limitations. Contingency theory models tend to be partial in that they focus on elements of control, such as budgeting, rather than addressing the full set of control devices. Agency theory models have similar weaknesses and also tend to focus on single period, single agent situations. As a result our understanding of MC structure remains fragmented. The purpose of this paper is to examine the potential of transaction cost economics (TCE) to provide a more useful way to explain management control structure variety.

2. An Introduction to TCE

TCE examines how different institutional arrangements are used to organize and control economic activity. Transactions differ in terms of contractual problems while organizational forms differ in their ability to solve related problems. TCE examines the match between transactions and organizational form while considering the bounded rationality (i.e., lack of cognitive and computational ability) and opportunism (i.e., "self-interest seeking with guile", p. 421) problems associated with human nature. 

The dimensions of transactions include: degree of asset specificity of the transaction, uncertainty, and frequency. Asset specificity is related to the opportunity losses incurred in cases where a transaction is prematurely terminated. The size of opportunity losses is influenced by information asymmetry and impactedness, e.g., where information related to a transaction is not evenly distributed between the parties to the transaction.

TCE defines three types of organizational structures used to govern transactions:
   1. Markets - control is derived from free competition.
   2. Hybrids - control is derived from long-term contracts.
   3. Hierarchies or internalization - control is derived from authority,
       internal structure and monitoring.

TCE's theme is that transactions tend to be matched or aligned with governance structures. Market control fits where transactions are low on asset specificity (i.e., low potential opportunity costs). Hybrid control fits better when asset specificity increases, but hierarchies are needed to reduce conflicts between contracting parties when transactions involve high asset specificity. Hierarchies support cooperation, but bureaucratic inertia, politicizing of decisions and other problems reveal that the hierarchical form of governance is not a panacea.

3. TCE and MC: Some Preliminary Considerations

According to Spekle, TCE has not had much influence on MC research. In this section, he examines the relationship between TCE and MC. Various definitions of MC are provided to show that MC includes two key concepts; a focus on behavior, and how this behavior effects organizational outcomes. He suggests that the lack of TCE's influence on MC may be related to the levels of analysis emphasized, where TCE emphasizes the tradeoffs between modes of governance, while MC emphasizes trade-offs within only one mode, i.e., hierarchy. He indicates that this is not an obstacle to TCE's application to MC. He goes on to point out the behavioral nature of TCE with its notions of bounded rationality and opportunism. This is followed by a discussion of some issues related to organization goals and effectiveness. For example, can organizations have goals and objectives? Or do the goals and objectives belong to  individuals associated with the organization? Where do these goals and objectives come from? One way to deal with these questions is to assume that the organizational goals are translated into strategies and the function of MC is to ensure that these strategies are implemented. But this approach ignores MC's role in crafting strategy. More of this sort of discussion is provided in this subsection, i.e., 3(4).

In the last part of section 3, Spekle indicates that MC-theory should be able to shed light on the functioning of various configurations of control devices and that TCE provides a procedure to help in this area referred to as a "remediableness check." This terms is also mentioned on p. 422.

4. A Transaction Cost Theory of MC

This is a fairly long section (pp. 427-438), but it is summarized in Tables 1 and 2 on pp. 437 and 438. My adaptations of these tables appear below. A control archetype is defined on p. 427 "as a characteristic, discrete configuration of control devices that is descriptively and theoretically representative of a significant group of observable management control structures and practices." Three variables define the activities to be controlled: Certainty/uncertainty, degree of asset specificity, and intensity of post hoc information impactedness. These variables are used to explain the various control archetypes as illustrated in Tables 1 and 2.

Certainty/uncertainty relates to whether activities are programmable or non-programmable. Programmable activities are those where the organization has sufficient knowledge related to how the activities should be performed along with the expected outcomes. Control for programmable activities (high certainty) can focus on compliance to specific rules and targets. Control over non-programmable activities must rely on a broad view of performance with much less precision. 

Market Control

Low asset specificity (idiosyncrasy) means that the assets involved are not tailored to the organization and can be efficiently controlled by the market pricing mechanism, i.e., Adam Smith's "invisible hand."

Machine Control

Programmable activities with high asset specificity (i.e., high potential for opportunity losses) tend to be controlled by standardization and regulation of behavior, measurement and monitoring of predefined outcomes. Speckle refers to this as machine control which has two sub-categories including action oriented and result oriented as indicated in the tables. Action oriented machine control involves rules and instructions related to a specific sequence of tasks followed by supervision/monitoring of actions. Results oriented machine control includes targets, accountability and rewards.

Arm's Length Control 

Arm's length control tends to be applicable where activities are programmable (certainty is high), but asset specificity or idiosyncrasy is moderate (i.e., moderate potential for opportunity losses). Arm's length control can be either hierarchical or hybrid (i.e., including hostage arrangements) as shown in the tables and is applicable where competition is thin and the contributor has significant autonomy. Hybrid arm's length control is involved in full outsourcing. An example of a hostage arrangement is where a contract includes a penalty payment for non-performance. Hierarchical arm's length control refers to situations where departments or divisions are treated as independent autonomous units, but management intervention is available when needed as in transfer pricing disputes.

Exploratory Control

Exploratory control is related to situations where information impactedness is low ( i.e., where the quality of a contributor's performance is clear to all). Exploratory control can be hierarchical or hybrid. Hierarchical exploratory control is applicable where activities have low programmability (the organization cannot define what it expects), asset specificity is moderate or high (, and information impactedness is low or remediable. Hybrid exploratory control provides an alternative with more market incentives and is applicable where activities have low programmability (the organization cannot define what it expects), asset specificity is high (i.e., high potential for opportunity losses), and information impactedness is low. Speckle indicates this archetype is applicable in outsourcing arrangements with a limited number of suppliers and depends on sharing private information and know-how.

Boundary Control

Boundary control is applicable where information impactedness is high (i.e., the quality of performance is not clear to other members of the organization). Boundary control refers to prevention of undesired actions or outcomes. Boundary control can be hierarchical or market-based. Hierarchical boundary control specifies behavior boundaries and is a structure of last resort. Market-based boundary control relates to situations where services are purchased in the external market, e.g., legal services or audit services.

5. Summary and Conclusion

The transaction cost theory (TCE) of management control (MC) provides a fairly comprehensive framework in terms of control archetypes that can be used to address any sort of control issue and it is empirically testable - at least in principle.

See the Simons Levers of Control summary.

	Web maaw.info