Mayer & Salomon 2006
Abstract
Independent and joint effects of contractual hazards and technological capabilities on governance
> Strong technological capabilities improve a firm's ability to govern transactions (making outsourcing feasible despite certain contractual hazards)
Sample: 205 service contracts from a single information technology firm
1) Contractual hazards encouraged internalizing transactions
2-1) Weak technological capabilities increase the likelihood of subcontracting
2-2) Strong technological capabilities had no independent effect
2-2-1) Only in the presence of certain contractual hazards, it had an impact
Explains why firms facing similar levels of contracting hazards organize their transactions differently
# With capabilities # Agency
* Introduction
Contracting hazards' key role in governance (Shelanski & Klein 1995)
The potential for capabilities to influence governance (Argyres, 1996; Leiblein & Miller, 2003; Mayer, 2006; Nickerson & Silverman, 2003; Nickerson & Zenger, 2002; Silverman, 1999)
How the resource-based view can complement the standard TCE approach to governance
A firm's technological capabilities (or the lack thereof) represent an important consideration when the firm makes governance decisions (Hoetker, 2005; Kogut & Zander, 1992; Leiblein & Miller, 2003; Martin & Salomon, 2003a, 2003b)
When transaction characteristics held constant, firms will outsource transactions when technological capabilities are weak and govern transactions internally when technological capabilities are strong
Capabilities come in many forms and variations technological, managerial, operational, marketing-based, etc.
Technological capabilities: superior ability to create a good or service that meets a customer's requirements for functionality, quality, and cost and schedule (Hoetker, 2005: 78).
1) Technological capabilities have been shown to have a strong influence on governance (e.g., Hoetker, 2005; Leiblein & Miller, 2003).
2) Technological capabilities are central to resource-based theories of firm-specific advantage (see Kogut & Zander, 1992; Mahoney & Pandian, 1992; Martin & Salomon, 2003b).
3) Technological capabilities have been shown to be particularly important in high-technology industries such as information technology, which is examined here (e.g., Mayer, 2006; Mayer & Argyres, 2004; Martin & Salomon, 2003b).
Hypothesis
Joint effects of contractual hazards and technological capabilities
RBV: strong technological capabilities -> integrate
Argument: technological capabilities can actually improve a firm's ability to govern interfirm, market transactions
Technological capabilities enable firms to select capable suppliers, effectively monitor their progress, and effectively share knowledge with them,
Then technological capabilities translate into "governance capabilities" that diminish costs imposed by contractual hazards
The ability to manage and monitor suppliers will play a larger role when transactions present "hold-up" and observability hazards, but less of a role when transactions are subject to appropriability hazards.
Sample
A random sample of 405 service contracts from a single firm between 1986 and 1998
The firm, a large information technology (IT) service provider, which we hereafter refer to as Compustar, provided access to the contract documents, each of which concerned a distinct project.
With the help of several engineers and managers at the firm, we were able to identify several capabilities and contracting hazards that formed the basis for measures of independent variables.
The importance of contractual hazards and firm capabilities in governance decisions (e.g., Leiblein & Miller, 2003),
-> None has explicitly examined such effects jointly.
Why firms facing similar levels of contractual hazards might rationally select different forms of governance
Capability differential -> The identification of "mistakes" and misalignments in governance choice
Firms develop governance capabilities that help mitigate contractual hazards and that these capabilities are a consequence of technological competence.
Governance capabilities help firms overcome some hazards but have little effect on others.
* Contractual hazards and capabilities in governance decisions
- Transaction cost economics and governance
1) "Holding up"
The extent of contractual hazards is shaped by the degree of asset specificity (Williamson, 1985), appropriability (Oxley, 1997; Pisano, 1990), or observability (Holmstrom, 1979) in an exchange # Opportunism
Asset specificity refers to the transferability of assets to alternative uses (Williamson, 1985).
2) Appropriability
Appropriability refers to contracting hazards that expose valuable intellectual property to expropriation (Gulati & Singh, 1998; Oxley, 1997; Pisano, 1990). #Leak
3) "Observability"
the observability (Holmstrom, 1979) When the quality of output is difficult to observe and measure ex-post, using market-based contracts to govern transactions can be problematic because it is unclear how to gauge, and thereby reward, quality. # Meterability
A firm may prefer administrative oversight so as to focus on controlling the input process rather than on metering the output (Holmstrom & Milgrom, 1991; Mayer & Nickerson, 2005) # Metering
Firms should internalize transactions when hold-up, observability, and appropriability concerns are great
Hypothesis 1. A firm is more likely to use internal forms of governance when a transaction is subject to contractual hazards.
- Capabilities and governance
Technological capabilities -> Governance decisions
Firm-specific capabilities are critical to a firm's success (Peteraf, 1993; Wemerfelt, 1984, 1995)
Distinctive technological, marketing, and managerial capabilities can be value creating for firms (Mahoney and Pandian 1992)
Technological capabilities impact governance decisions (e.g., Hoetker, 2005; Leiblein & Miller, 2003).
Why 'buy' in capabilities
One of the key assumptions associated with the resource-based view is that capabilities are inherently complex, causally ambiguous, and difficult to replicate (Barney, 1986, 1991; Conner & Prahalad, 1996; Kogut & Zander, 1992).
It is more efficient for the firm to use market forms of governance to gain access to the skills and capabilities that it lacks because it will be very difficult, costly, and time-consuming to try to develop those capabilities from scratch.
+ Uncertainty surrounding the "true" value of the capabilities results in firms overpaying for these assets when trying to acquire them in the open market (Dierickx & Cool, 1989). # 하지만 반대도?
+ Even if a firm internalizes transactions through acquisition, it may be unable to assimilate the complex capabilities of the acquired organization (Kogut & Zander, 1992; Martin & Salomon, 2003a).
Strong capabilities.
Because internal capabilities have a public-good character, exploiting them is more efficient within, rather than across, firm boundaries (Kogut & Zander, 1992; Martin & Salomon, 2003a).
For transactions drawing upon technological capabilities in which a firm has an advantage, the firm should prefer to internalize to leverage its existing capabilities.
Hypothesis 2a. A firm is more likely to use market forms of governance when a transaction draws upon a technological capability in which the firm is weak (i.e., a technological capability that the firm lacks)
Hypothesis 2b. A firm is more likely to use internal forms of governance when a transaction draws upon a technological capability in which the firm is strong (i.e., a technological capability that the firm possesses)
# Bring the variances from the firm capability, but in my opinion, the sources of each cost follow TCE's logic more. eg. production/governance cost
# Same as Pisano (1990)
- Governance capabilities
How firm capabilities moderate the effect of contractual hazards on governance mode
Firms may develop capabilities in governing certain types of transactions (e.g., Dyer & Singh, 1998)
Transactions -> Capability
Capacity from transaction
Firms may develop capabilities in governing certain types of transactions (e.g., Dyer & Singh, 1998)
Not only does the creation of technological knowledge through strong R&D matter to firm growth and profitability, but also that complementary technology transfer skills are a substantial source of competitive heterogeneity (Galbraith 1990)
According to Cohen and Levinthal (1990), one such complementary capability, absorptive capacity, results from investments in R&D.
Martin and Salomon (2003a) formalized technology transfer as a firm-specific capability.
Effect: decreases the costs of sharing knowledge and bestows upon a firm the capability to better protect proprietary assets during technology transfer
Author's logic (technological capabilities -> governance capabilities)
Just as R&D investments can play a role in the development of technology transfer capabilities, technological capabilities can play a role in the development of governance capabilities
Although strong technological capabilities certainly lower the cost of internal production, a firm with these capabilities may also, because it understands the technologies relevant to a project, be better able to identify appropriate project suppliers and avoid low-quality sub contractors in a potential "market for lemons" (Akerlof, 1970)
Overcome potential information asymmetries
1) better evaluate a partner's skills, judge its readiness to perform the task, assess its ability to accept and receive guidance, and provide such guidance through technology transfer when necessary.
2) strong technological capabilities lower the cost of external governance by enabling more effective monitoring
A firm with superior technological capabilities to be able to govern external market exchanges better than a firm without such capabilities.
Context
Technological capabilities are likely to be especially valuable when a given transaction is subject to contractual hazards, as the costs associated with the external governance of high-hazard transactions decrease in technological capabilities at a rate faster than internal production and governance costs (see Hennart, 1993).
Scenario
1) A firm chooses to use a supplier for a transaction that faces contractual hazards, the firm will try to craft a contract that provides safeguards to mitigate the hazard.
2) If the firm has strong technological capabil ities in an area germane to the contract, it will be better equipped to design the contract with the right amount of detail, contingency planning, and pecuniary incentives.
3) If the transaction were not subject to contractual hazards, we would expect the firm, regardless of its technological capability, to easily craft such a contract
Governance capabilities (How capabilities mitigate problems of contractual hazards)
"Hold-up" / Observariliability (++)
Appropriability (+?)
Some technology inevitably spills over when firms transact willingly with partners (Martin & Salomon, 2003a); and, as Shapiro and Varian (1999) pointed out, contracts are only minimally effective at preventing technology leaks.
# Hold-up: a firm makes an investment that is specific to a given transaction, while the contractor takes actions that devalues the initial firm's investment (Masten, 1996)
Hypothesis 3a. The relationship between contractual hazards arising from hold-up and internal governance decreases in the presence of
Hypothesis 3b. The relationship between contractual hazards arising from observability and internal governance decreases in the presence of strong capabilities.
Hypothesis 3c. The relationship between contractual hazards arising from appropriability and internal governance is unaffected by the presence of strong capabilities. # How to measure, Logics following the kinds of contractual hazards
* Data and methods
An industry characterized by technology-intensive products, and one in which firms have the option to outsource
-> Information technology services industry
1) The industry is characterized by numerous individual projects that firms can either outsource or complete in-house.
2) The industry is large in scale and scope, allowing us to observe a wide range of projects involving an assortment of hardware and software solutions and drawing upon varied technological capabilities.
3) we had access to all contracts for the IT projects performed by one firm, we could analyze individual contracts for contractual hazards. All
Transaction as our unit of analysis (e.g., Shelanski & Klein, 1995).
A project-based industry such as IT, the repeated exchange of an identical good is largely absent < Each project is a separate transaction
Focused on the project, as a self-contained transaction, as the unit of analysis.
The customization of service projects in many high-technology industries means that suppliers in these industries must make sourcing decisions? use in-house resources or use subcontractors on the basis of their technological capabilities as suppliers and an evaluation of the attributes of each project.
Capabilities in this industry, although certainly specific to firms, are generally developed around given technologies tailored to particular hardware, software applications, and programming languages.
Once a supplier is awarded a project, the supplier must decide how to fulfill the project requirements? sometimes using internal resources, sometimes outsourcing work to subcontractors.
Data on IT service projects are undertaken by one firm, Compustar. (Compustar, a producer of mainframes and related hardware since the 1970s, entered the platform-independent IT services business in the mid-1980s)
We randomly selected a sample of 405 of Compustar's IT contracts with U.S. customers, a figure representing approximately 25 percent of the entire population of contracts
Our final sample, therefore, included detailed information at both the firm level and the individual contract level for 405 of Compustar's IT service contracts over the period 1986-98.
- Sample
SUBCONTRACT: Our dependent variable captured whether Compustar governs a particular transaction internally or elects to use an external subcontractor (1: contract, 0: in-house)
- Measures
Contractual hazards
INTERDEPENDENCY: a proxy for "hold up"_Interdependence was coded 1 if a customer's personnel were directly involved in the project in such a way that Compustar had to depend on them to complete the task(s), and 0 otherwise. #?
TIMESPECIFICITY: a proxy for "hold up"_a dichotomous variable coded 1 if such a spcified date existed in the contract and 0 otherwise. #? (Masten, Meehan, & Snyder, 1991; Pirrong, 1993) Empirical tradition.
DISRUPT: a proxy for "hold up"_a project could shut down a "significant portion" of a customer's data center and as 0 otherwise.
LACKOBSERVE: a proxy for "observarability"_coded 1 if the quality of the end product associated with a particular project was difficult to observe and 0 otherwise
APPROPRIABILITY: a proxy for "appropriability"_Proprietary technologies subjected valuable knowledge to the risk of misappropriation by subcontractor. If any of these technologies or processes were to be used for a transaction and as 0 otherwise
Technological capabilities
Capabilities were identified through detailed discussions and interviews with Compustar employees and manag ers and were verified by engineers from other firms in the IT services industry.
HARDWARE: Advantageous capabilities_coded 1 if a project involved working on Compustar-manufactured hardware and 0 otherwise.
MAINFRAME: Advantageous capabilities_coded 1 if a contract involved working on a mainframe computer and 0 otherwise.
PROGRAM: Not advantageous capabilities_took the value 1 if a project primarily involved programming tasks, 0 otherwise
OTHERWARE: Not advantageous capabilities_coded 1 if a project involved work ing on hardware from another vendor and 0 other wise
Measures of capabilities are consistent with others in the literature (e.g., Mayer & Nicker
son, 2005) and with the more general experience based measures of capabilities used in earlier stud ies (e.g., Leiblein & Miller, 2003)
# Firms may win business even without superior technological capabilities for three reasons.
1) a strong relationship with the customer may have given the firm an inside track on the project.
2) Compustar may have a reputation for quality, which makes the de cision to use them easier to justify to a board of directors, especially if there is a problem.
3) Compustar may have submitted a lower bid than competitors with superior capabilities. A
Controls
NUM: The number of previous projects that Compustar had performed for a specific customer.
Compustar personnel indicated that they liked to do the first few projects with new customers in-house because they believed that using in-house resources helped them develop a closer relationship with the customers and increased the potential for future business.
log(SIZE): The sheer size of a project. (as the total dollar value of a project's revenue, expressed as a natural logarithm.)
Large projects were more likely to exceed the capacity of single suppliers and thus required outsourcing.
<-> However, large projects might also have been the ones that were more critical to Compustar projects that they might have preferred to perform in-house.
FREE: "Unabsorbed slack"_A greater amount of cash implies that a firm has more than enough resources to hire personnel to perform projects in-house (see Tan & Peng, 2003)_the natural log of Compustar's available cash (in millions of U.S. dollars) in a given year
A greater amount of cash implies that a firm has more than enough resources to hire personnel to perform projects in-house (see Tan & Peng, 2003).
PEAK: "Absorbed slack"_a firm's human resource capacity constraints (Tan & Peng, 2003)_defined peak months as a dummy variable coded 1 if a project originated during one of the four peak months and 0 otherwise.
Firms with substantial slack resources have the personnel available to perform work in-house even during peak months if they wish. O
Measures of organizational slack, defined by Bourgeois as "a cushion of actual or potential resources which allow an organization to adapt successfully to internal pressures for adjustment or to external pressures for change in policy, as well as to initiate changes in strategy with respect to the external environment" (1981: 30). # Slack is important, - Entrepreneur
Slack resources are those a firm has in excess, in amounts over and above those required for survival, and that allows it to pursue growth strategies (Cyert & March, 1963).
Our treatment of slack is consistent with other prevailing measures (e.g., Bromiley, 1991; Reuer & Leiblein, 2000; Tan & Peng, 2003).10
TIME: time fixed effect
They indicated that Compustar initially subcontracted extensively, as demand was growing faster than Compustar's internal capacity.
- Statistical method
We employed the probit model because its maximum-likelihood estimation procedure is particularly appropriate for dealing with qualitative data of the sort collected in this study (Maddala, 1983).
* Results
Base (M1), Contractual hazard (M2), Technical capacities (M3), Full model (M4)
Bigger size -> Subcontract: Exceed the capacity
Interdependency (Hold up) -> In-house: Specific to customer
Appropriability -> In-house: Leakage concern
Disrupt (Hold up) -> In-house: Important to customer
Lack of observability -> In-house: hard to control subcontractors
Programming -> Contract: Not advantageous
Other hardware -> Contract: Not advantageous
Insignificant
Previous projects
Free cash flow
Peak months
Time specificity
Computer hardware
Mainframe
Interaction model
Strong_Index of 2 (-) 1.22: In-house
Hold up_Index of 3 (-) 0.50: In-house
Strong * Hold (+) 0.64: Contract
Compared to Weak + Unspecific
Strong + Unspecific: In-house
Weak + Specific: In-house
Compared to Strong + Unspecific
Strong + Specific: Contract (Interesting): Governance capability
Compared to Weak + Specific
Strong + Specific: less In-house: Governance capability
Observability (X)
Appropriability (X)
No interaction effect of weak capabilities
* Discussion and conclusion
Rather, therefore, than ask the question "What is the best generic mode (market, hybrid, firm, or bureau) to organize X?" which is the traditional transaction cost query, the question to be put instead is "How should firm A?which has pre-existing strengths and weaknesses (core competencies and disabilities organize X?" (1999: 1103)
These findings support the view that strong technological capabilities make it easier for a firm to identify, monitor, contract with, and man age contractors.
Technological capabilities affected the governance decision only in the presence of hold-up hazards, not in the presence of observability or appropriability hazards.
1) they highlight the need for a fruitful integration of transaction costs and resource-based perspectives that carefully addresses how different capabilities affect different contractual hazards
2) these findings also highlight the limitations of governance capabilities derived from technological capabilities
Highlight the contingent nature of governance decisions.
Governance capabilities
Dyer and Singh (1998) introduced the concept of interorganizational advantage and the closely related concept of alliance capability (e.g., Kale, Dyer, & Singh, 2002).
Further, our research raises the possibility that governance capabilities may be a more general capability that can be leveraged across all sorts of interorganizational relationships (alliances, joint ventures, etc.).
Williamson (1985, 1996) argued that contractual hazards were a primary driver of governance decisions; however, left unexplained was why two firms facing transactions with identical levels of contracting hazards would select different governance structures. Sc
We did not explicitly test the performance implications of Compustar's governance decisions
The extent that we have uncovered a general governance capability, future studies could examine how such capabilities develop. For
What role does organizational culture play in the development of governance capabilities?
First, this study relies on microanalytic data from a single firm.
Second, we recognize that our measures of firm capabilities are inherently subjective.
Third, in this paper we focused on a specific governance capability arising from technological capabilities. W
What point do strong hazards or weak capabilities dominate when they are present in the same transaction?