Knowledge, Market Failure and the Multinational Enterprise: A Comment
- PDF / 2,524,564 Bytes
- 7 Pages / 450 x 771 pts Page_size
- 22 Downloads / 199 Views
*DonaldG. McFetridgeis ProfessorandChairin the Departmentof Economics, Carleton University,Ottawa,Canada.He specializes in IndustrialOrganizationandhas publisheda numberof papers on both multinationalenterpriseand the economics of technological change. Received:November 1994; Revised:January1995;Accepted:January1995. 409
Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve, and extend access to Journal of International Business Studies ® www.jstor.org
410
JOURNALOF INTERNATIONALBUSINESSSTUDIES,SECONDQUARTER1995
A numberof empiricalstudieshave foundthatthe probabilitythata transactionis internalizedincreaseswith the importanceof transaction-specifichumancapital. These studies include Monteverdeand Teece [1982], Andersonand Schmittlein [1984] andMasten,Meehanand Snyder[1989]. Love could have gone further.Complexity aggravatesan asset specificity problem. Given the level of transaction-specific investment, greater complexity increasesboth the frequencywith which contractualadaptationmay be required and the extent to which opportunismcan be disguised. In the simplest terms, when more things can go wrong, it is easier to find a pretextfor a self-interested alterationof one's contractualobligations. This is called shock or signal distortion [Wiggins 1990]. Complexity implies both more opportunitesand a greater abilityto alterthe termsof tradeex post. The role of complexity is fairly well establishedin the empiricaltransactioncost literature.Masten,Meehanand Snyder [1991] find that,given temporal,physical capital and humancapital specificity, the probabilityof internalizationin naval shipyardprocurementis an increasingfunctionof taskcomplexityfor high levels of complexity.Masten[1984] finds that,given site anddedicatedasset specificity, the probabilityof internalizationin aerospace procurementincreases with the technicalcomplexityof the product. K&Zcorrectlynote thatcomplexity has been measuredin manyways. It is transactionalcomplexity thatis relevant.Technologicalcomplexityor taskcomplexity may or may not imply transactionalcomplexity.This raises the usual problemof testing the data and the hypothesis simultaneously.As more empirical work is done, it may be possible to determine the robustnessof empirical results with respectto alternativemeasuresof complexity.The body of empiricalworkreferred to below implies, for example, that the influence of humancapital specificity is morerobustto differentmeasuresof thatvariableanddifferentspecificationsof the transactioncost modelthanis the effect of site specificity. K&Zalso arguethatcomplexityneed not imply a hold-upproblem.This is correct. Complexityaggravatesthe ex post small numbersor lock-in problemcaused by transaction-specificinvestments.If therewere no transaction-specificinvestments hence no lock-in, complexity need not burdenmarketand unified governance differentially. K&Z go on to make a differentand, in my view, incorrectargument.They argue thattherecan be no hold-upproblem in theircase becauseall the technologies in theirsample face comp
Data Loading...
