View source for Williamson (1991) - Comparative Economic Organization The Analysis Of Discrete Structural Alternatives

{{Article
|Has page=
|Has title=
|Has author=
|Has year=
|In journal=
|In volume=
|In number=
|Has pages=
|Has publisher=
}}
*This page is referenced in [[BPP Field Exam Papers]]


==Reference(s)==

Williamson, Oliver (1991), "Comparative Economic Organization: The Analysis of Discrete Structural Alternatives", Administrative Science Quarterly, Vol. 36, No. 2. pp. 269-296 [http://www.edegan.com/pdfs/Williamson%20(1991)%20-%20Comparative%20Economic%20Organization%20The%20Analysis%20of%20Discrete%20Structural%20Alternatives.pdf pdf]

 @article{williamson1991comparative,
  title={Comparative economic organization: The analysis of discrete structural alternatives},
  author={Williamson, O.E.},
  journal={Administrative science quarterly},
  pages={269--296},
  year={1991},
  publisher={JSTOR}
 }

==Abstract==

This paper combines institutional economics with as- pects of contract law and organization theory to identify and explicate the key differences that distinguish three generic forms of economic organization-market, hybrid, and hierarchy. The analysis shows that the three generic forms are distinguished by different coordinating and control mechanisms and by different abilities to adapt to disturbances. Also, each generic form is supported and defined by a distinctive type of contract law. The cost- effective choice of organization form is shown to vary systematically with the attributes of transactions. The paper unifies two hitherto disjunct areas of institutional economics - the institutional environment and the institutions of governance - by treating the institutional environment as a locus of parameters, changes in which parameters bring about shifts in the comparative costs of governance. Changes in property rights, contract law, reputation effects, and uncertainty are investigated.


==Discrete Structural Analysis==

There is not a continuum of organizational forms:
*Firms are not extension of markets but employ different means
*Discrete contract law differences support and define each form
*Marginal analysis is concerned with second order effects - economizing is first order

The technology view suggest that no novel economizing issues occur within firms (i.e. there are still economies of scale and scope), and that firms are just instruments for converting inputs into outputs. In contrast, Williamson claims that:

 "hierarch is not just a contractual act but is also a contractual instrument,
               a continuation of market relations by other means"
                                                             -Williamson, 1991

Also, note that the hybrid form discussed here is not a loose amalgam of market and hierarchy, but posesses its own disciplined rationale.


==Contract Law==

Williamson again puts forward forms of contract law:
*Classical Contract Law (covered in [[Williamson (1979) - Transaction Cost Economics | Williamson (1979)]])
*NeoClassical Contract Law (new material covered below)


===NeoClassical Contract Law===

Is characterized by the excuse doctrine and forebearance

Key characteristics are:
*Parties maintain autonomy but are bilaterally dependent
*Continuity of the relationship and adaptation are stressed
*Hybrid modes of contracting are supported
*There is mediation:
**For example in regulation, the regulator mediates
**Arbitrators
*Franchising requires added supports (because of mediation problems)


There are three kinds of disturbances:
*inconsequential - deviations are too small to recover costs of adjustment
*consequential (mid range)
*highly consequential


For consequential adjustments, the contract is a framework, providing guidance and norms. Contracts which support consequential adjustments:
*explicitly contemplate unanticipated disturbances which require adaptation
*provide tolerance zones in which misalignments will be absorbed
*require information disclosure and substantiation if an adaptation is needed
*provide for arbitration if voluntary agreement fails.


Again, arbritation has advantages (ease of education, etc), but if the disturbance is too large, arbitration may fail and result in litigation. Here there an excuse doctrine to appeal to, for example:
*impossibility
*frustration
*mistake
*manipulative interpretation
*duress
*undue influence
*bankruptcy laws
*and so forth.


Therefore there is a trade off between incentives and reduced opportunism. If the disturbance was truely unforeseen then a strong excuse doctrine reduces the gains to opportunism (while maintaining performance incentives). If excuse is routinely granted, then planning to avoid aversity will be reduced.


Arbitration is very costly, and the range of disturbances it can support is limited. As disturbances get very large, this framework is no longer optimal.

==Forebearance==

The firm is a "nexus of contracts", but with importance differences. Specifically it has fiat (the power to order something to be done). This may have its origins in the employment contract, and/or that of forebearance. Forebearance is the court's unwillingness to hear disputes between internal divisions of a firm. 

       "Hierarchy is its own court of ultimate appeal"
                                    -Williamson, 1991
                                    


The "business judgement rule" holds absent bad faith - directors are not liable for their corporation's mistakes. There is a quasi-jurisdicational barrier.

The court still hears worker's rights complaints etc, but does not interfer in the operation of firms generally. 


Unions may refuse to bring individual grievences to arbitration as this:
*Discourages day-to-day cooperation
*The claim may endanger group interests
*The union is an arbitrator of a relationship with acceptable bounds...


To get fiat, internal incentives must be flat - so that workers incentives do not conflict with their orders. 


The underlying rationale for forebearance is:
*parties have deep knowledge of dispute and efficient solutions (it would be costly to communicate them, and they may not be verifyable outside the organisation)
*Permitting internal disputes in courts would undermine the nature of the firm.


==First Order Economizing==

First-order economizing is effective adaptation and elimination of waste. Economics was (is) too preoccupied with issues of allocative efficiency (marginal analysis is used), to the neglect of organizational efficiency (i.e. the choice of organizational form).


==Dimensional Governance==

The discrimating alignment hypothesis is that transactions (which differ in their content) are aligned with governance structures (which differ in their costs and competencies). Market, hybrid and hierarchy differ in contract law respects, but there is more to governance than just contract law. Adaptability and incentive/control instruments are key.


===Adaptation===

Economic problems arise as a consequence of change. The price system is an efficient mechanism (in the market) for communicating information and inducing change (Hayek 1945). However, firms also adapt to change (Barnard 1938), and concious, deliberate, and purposeful cooperation is realized through formal organization (firms). When the price is a sufficient statistic, then the market excels. The sort of adaptation that are supported through price signals are autonomous (<math>A\,</math>) changes. However, some disturbances require coordinated (in later papers refered to as cooperative) responses, lest the components operate at cross purposes. These are <math>C\,</math> changes. Problems in addressing <math>C\,</math> changes include:
*If one party coordinates, then they may use their position strategically for their own gain
**They may distort information, or disclose it selectively or incompletely.
*There may be self-interested bargaining (which is costly)
*Maladaption itself is costly.

Fiat has advantages for <math>C\,</math> changes.

===Instruments===

With <math>A\,</math> changes, incentives are aligned by prices in markets. They are also high-powered. Autonomous traders can't make claims against other autonomous traders.


In contrast <math>C\,</math> changes give advantages to hierarchies, but this comes at some cost. Divisions (or individuals) can misreport information to effect strategic redistributions. Internal organization degrades incentive intensity, which results in added bureaucratic costs (note that incentive intensity may be deliberately suppressed to prevent it getting in the way of bilateral adaptation). There is also a differential reliance on controls. To offset the loss of incentives, firms need greater controls:
*monitoring
*career rewards
*penalties


===Hybrids===

Markets and hierarchies are extreme forms. In the middle are hybrids. They are supported by neo-classical contract law, and display intermediate values with respect to:
*<math>A\,</math> changes - ownership autonomy is preseved giving strong incentives.
*<math>C\,</math> changes - bilateral dependency adds safegaurds (information disclosure, dispute-settlement apparatus)
**But these dull the incentives
**And the are more costly than in firms (proposals need more documentation, arbitration is more costly than fiat, information is harder to access and validate, there are less instruments to promote team orientation)
**Extreme changes will break the organizational form
*Contract incompletness - martkets have complete, firms have very incomplete
*Adminstrative apparatus (see above).


==Discriminating Alignment==

Discriminating alignment uses both governance costs (detailed above) and asset specificity arguments.


===Asset Specificity===

 Asset specificity is the degree to which an asset can [not] 
 be redeployed to alternative uses and by alternative users 
 without sacrifice of productive value. 
                                           -Williamson, 1991

Six distinctions are made:
#Site specificity (to economize on inventory and transportation)
#Physcial asset specificity (custom dies etc)
#Human asset specificity (learning by doing)
#Brand name capital
#Dedicated assets - discrete investments made in a general purpose plant for a particular customer
#Temporal specificity - technological non-seperability (e.g. timely response of humans)

The first five forms (in particular) create bilateral dependency.

===Alignment===

Ignoring the revenue and production-cost consequences of specificity, one can plot the governance costs as a function of the asset specificity. 


Hybrid forms, sacrifice incentives but assure that the parts are better at not working at cross-purposes. Selling in the market gets full incentives, but free-riding on the brand name and other suboptimization may result. Franchising gives greater autonomy than hierarchy but addes rule and surveillance as compared with markets: Suboptimizating is reduced relative to markets, and incentives are sharpened. Though not that local autonomy may get in the way of global adjustments and vice versa. For example, franchisees can not reduce costs by using local materials if this would undermine the global brand. 


==Comparative statics==

Noting that the choices are discrete, the paper provides some 'comparative statics' with respect to property rights, expropriation, contract law, reputation effects and uncertainty.


===Property Rights===

Property rights consist of three elements:
#The right to use the asset
#The right to appropriate returns from the asset
#The right to change the asset's form and/or substance


There are two security hazards associated with property rights: expropriation by government and expropriation by commerce (rivals, suppliers, customers).


====Expropriation by Government====

Issues of credible committment are pertinant. If property rights are subject to occasional reassignment (and full compensation is not paid at each instance), then strategic considerations include:
*Reallocating wealth (investing in things that can not be expropriated like consumption) away from long-term (non-moveable) assets
*Incentive to adapt are muted


The government will have to bear a larger durable investment burden. Note that is can be difficult to commit. Previous renaging may prevent future committments from being believed. The king (or authority) must be subject the the same law as the everyone else, and committment involves changing the institutional environment in a way so that it can not be unchanged again.


====Leakage====

Value may be appropriated or disappated by rivals (etc). Weak appropriability regimes (c.f. [[Teece (1986) - Profiting From Technological Innovation | Teece (1986)]]) mean that:
*Ex ante investment incentives are impaired
*Ex post incentives to embed investments in protective governance structures are impaired. 

Vertical or lateral integration offers protection. 


===Contract Law===

If excuse doctorine were to become too strict or too lax this would affect investment incentives. This was covered above.


===Reputation Effects===

Non-heirarchical contracts may use reputation effects to lower opportunism. This would increase the use of hybrid contracting relative to hierarchy. Note that ethnic communities with a high degree of solidarity often enjoy hybrid contracting advantages.


===Uncertainty===

Greater uncertainty could come from more disturbances or more important disturbances. The hybrid mode is most susceptible to such an increase. Because adaptations can not be made unilaterally, they take time (and cost more). At some level the disturbances could come so fast that the next arrives before the last has been dealt with. Thus the hybrid mode may be infeasible at very high levels of uncertainty. These "timelines" arguments also impact R&D joint ventures, unless they are especially designed to support quick responsiveness. 


===Simultaneous Parameter Shifts===

The paper points out that using policy to address a single factor in isolation may fail. Suppose that privatization reform required banking and property rights together, then introducing only banking policy would have no effect.