Organizations and Markets

20 Comments Add your own

• 1. Martin  |  6 February 2014 at 2:07 pm

  Professor Klein,

  could you please elaborate more on what is the current Austrian stand on the concept of transaction costs? I remember an article by Rothbard where he rejects the idea of transaction costs since all costs are purely subjective. I think a similar point was made by Guido Hulsmann in his excellent paper “The Apriori foundations of property economics” (if I recall the title correctly).

  Thanks

• 2. Peter Klein  |  6 February 2014 at 3:05 pm

  Rothbard (e.g., here) doesn’t reject the concept of transaction costs per se, but he says that transaction costs — in the “neoclassical” sense described in my post — are subjective, just like any other kind of cost. His main target here is the related concepts of “social cost,” and “social efficiency,” and the kind of utilitarian, minimizing-society’s-transaction-costs approach one typically finds in Chicago law and economics. By the way Rothbard embraced Coase’s theory of the firm in his own work on vertical integration (see the quote on p. 17 here). Huelsmann likewise doesn’t reject the concept of (subjectively understood) transaction costs, but he thinks the concept — again, in the first of the two senses mentioned above — is not particularly useful. You might also look at Roy Cordato’s work on efficiency.

• 3. Joe Mahoney  |  6 February 2014 at 3:43 pm

  Kenneth Arrow defined transaction costs as “the costs of running the economic system” (1969, p.48). Such costs are to be distinguished from production costs, which is the cost category emphasized in neoclassical economics, as you noted Peter. Transaction costs are the economic equivalent of frictions in physical systems.

  Transaction costs of ex ante and ex post are usefully distinguished. The first are the costs of negotiating, drafting, and safeguarding an agreement. The ex post costs can also take several forms including maladaptation costs to which you refer Peter, but could also include haggling costs to correct ex post misalignments, as well as the setup and running costs associated with governance structures (typically not the courts) to settle disputes.

  Addressing another point, Peter, I would fold in the Jensen and Meckling (1976) agency costs of monitoring costs, bonding costs, and residual loss as a subset of transaction costs.

  You raise an interesting question about dropping the term transaction costs and considering only more precise terms. The reason to be cautious about this solution is that it may direct attention away from considering both ex ante and ex post transaction costs simultaneously. Ex ante and ex post transaction costs are interdependent and must be addressed simultaneously rather than sequentially.

  At a more philosophical level, is there not vagueness inherent in all concepts, such as when psychologists talk about “intelligence” as they “mismeasure man” (Gould, 1981). Judging whether a concept is precise in an absolute sense is mis-specified. All concepts are indefinitely indefinite. Thinking in terms of a “comparative assessment of imperfect alternatives,” in what sense is the concept of “transaction costs” less precise than other concepts in social science, and by what criteria?

  Finally, pragmatically speaking, the transactions costs that need to be emphasized in a particular paper depend upon the problem at hand.

• 4. michaelfto  |  6 February 2014 at 8:34 pm

  I will have to read the cited papers.

  When I was in Law School, after doing PhD in Choice Theory, we were introduced to Coase’s theorem in a Law and Economics class.

  The first thing I asked for was a definition of “transaction costs”. The theorem looked remarkable thin to me with a definition.

  Guess after 20 years, I can now learn something. Thanks!

• 5. Sanjeev Sabhlok  |  6 February 2014 at 10:58 pm

  I agree that the concept of transaction costs, while extremely valuable, is loose and ill-defined.

  In any transaction or action (and all economic exchange is a transaction or action), all costs need to be identified and counted. So long as we count all costs, we can’t really go wrong. We can call them what we like, but we should define them precisely.

  In general, I prefer the much clearer approach of Demsetz which rejects transaction costs as a ‘category’ and thus rejects the unnecessary qualification made by Coase.

  We could (only as an example) break up the costs of a transaction into direct and indirect costs. Direct costs could then be defined as those paid through cash (or equivalent), with indirect costs being largely the opportunity costs of the incremental time involved from the moment one starts thinking about a transaction up to the moment it is completed. At each stage the costs of such time (and often cash) spent would be evaluated on the opportunity cost principle, with only time that falls on a critical path being counted.

  There are many ways of counting costs. The Regulatory Change Measurement manual of Victoria (Australia) measures the costs imposed by regulation and classifies into administrative costs, compliance costs delay costs, financial costs, and indirect economic costs. Precise definition allows for precise measurement. That, I think, is the key.

• 6. Jim Rose  |  7 February 2014 at 2:46 am

  see http://www.yorambarzel.com/ for barzel’s asute writings on waht are transaction costs.

  see barzel’s transactions costs: are they just costs?

  in his book barzel defines transaction costs as the costs associated with the transfer, capture, and protection of rights.

• 7. Aidan Walsh  |  7 February 2014 at 10:28 am

  I don’t understand transactions costs as part of the theory of the firm.

  Markets for property, clothes, investments, education, funding, books, music, spouses and everything else have been transformed by the internet. Information flows and transaction costs have been utterly changed. Even the cost of contracting have reduced, which we can appreciate every-time we click on ‘I Accept’. But none of this has made much difference to the firm. Amazon.com does something quite new, but Amazon the firm is run on a basis that would be entirely familiar to, and in certain ways vindicate the theories of FW Taylor.

  For Hayek, rules were central

  So how would a rule perspective help us develop a different rationale why firms have emerged from the market process? As Hayek pointed out, there IS a problem with the extended market order and that problem is unaffected (or perhaps even made worse) by changes in technology: the rules of the extended order must be common throughout and those rules must be ‘of a somewhat smaller content than one applied to a small group’ (Hayek, 1976, p. 89). If individuals make decisions by following rules ( or ‘heuristics’ if you will, Gigerenzer, Kahneman), and groups of individuals make coordinated decisions by following common rules, then an extended market order based on common attenuated rules restricts the set of possible different decisions that can be made and thus restricts the range of solutions available (Casson, 1995, p. 89). Firms, each with their own individual body of rules, different rules from each other and from the rules in the larger market order, can be seen as the solution to this problem. In this light we can see the firm, first coming to prominence as large-scale markets developed in the 19th century (Chandler, 1977, p. 498) as the market throwing up its own solution to the problem-solving problem created by common attenuated rules across an extended market order.

  Who needs transaction costs?

• 8. Jon Bingen Sande  |  7 February 2014 at 10:32 am

  When I explain these issues to my students, I use Gibbon’s (2005) framework as a starting point, with 1) a contract negotiation stage, 2) a value creation stage (ex ante efforts and investments), 3) a change in the state of the world stage, 4) a value claiming stage (ex post decisions), and 5) a payoff stage. Then I explain where three different problems might arise (measurement problems, adaptation problems, and safeguarding problems) and how these problems lead to different kinds of costs.

  These costs seem to largely overlap with those that Joe describes:
  – Ex ante direct transaction costs of search, negotiation, and monitoring
  – Ex ante opportunity cost of efforts and investments not made (i.e., lost values from ex ante efforts and investments )
  – Ex post direct costs of enforcing agreements (haggling, renegotiations etc).
  – Ex post opportunity cost due to maladaptation (i.e., lost values due to failure of adaptation to new situations)

  I usually find it difficult to explain the term opportunity costs – but some authors in the marketing literature do, like Rindfleisch and Heide (1997) and Ghosh and John (1999) (I teach B2B marketing). Therefore, I usually try to explain that we’re really here talking about a loss of value, or lost opportunities to create value because a poor governance choice has been made. For example, choosing a fixed-price contract for a complex development project where ex post adaptations should be made, will often prevent the parties from doing the ex post adjustments that they ideally should do if they want to maximize joint value.

  However, I never feel 100% sure this is the best way to explain it…

• 9. Joe Mahoney  |  7 February 2014 at 11:10 am

  Thank you Peter for revisiting this question. There is much to learn from these posts, which is blogging at its best. Structuring a doctoral seminar session on transaction costs along the lines of Professor Jon Sande’s outline has merit. Other ideas to try to do even better will, of course, be much appreciated.

• 10. Peter Klein  |  7 February 2014 at 12:19 pm

  Yes, this a very helpful discussion. I should perhaps have mentioned that Doug Allen is a Barzel student and his “property-rights” definition of TCs is based on Barzel’s ideas.

  I appreciate Aidan raising Hayek’s point about rules — I have reviewed or discussed several papers in recent years on relational contracting and other kinds of informal arrangements, and have tried to point the authors towards Hayek and the general field of constitutional economics. “Framework contracts,” for example, are similar to constitutions. But I’m not as convinced as Aidan that this approach answers the basic Coase (1937) questions. (And on the point about the internet, remember that IT also lowers the costs of internal coordination — it reduces “internal transaction costs” as well as “external transaction costs” — so the net effect on firm size is ambiguous, ex ante. See Varian: http://www.nytimes.com/2002/01/17/business/17SCEN.html.)

• 11. Randy  |  7 February 2014 at 3:58 pm

  Looks like time for a serious workshop on the social ontology of TC. Perhaps John Searle could keynote… Finding the shared social meaning of the terms and categorizations would be a triumph of scholarship in our field (however defined). Thence, on to putting real meaning to “entrepreneurial opportunity” and “routines”. Thanks for dropping the puck, Peter.

• 12. Massimiliano Vatiero  |  8 February 2014 at 5:13 am

  Dear all,

  I agree on the fact that there is not an explicit and clear definition of transaction costs. Very often (see Coase 1937, 1960) there is a list of transaction costs. Moreover, Do transaction costs represent opportunity costs?

  Finally, transaction costs may include also transiction/switching costs, namely costs for “moving” from an institution (for instance the market) to another (for instance, the firm). They represent the costs of “institutional reforms”. In the coasian argument they seem null or irrelevant.

  More generally, in my opinion the crucial problem is that there is not a clear definition of transaction. Williamson advocated John Commons’s notion of transaction, but he follows this notion only partially.

• 13. Jim Rose  |  8 February 2014 at 5:28 am

  I should add that barzel defines the firm by its guarantee capital and by the scope of its guarantees.

  The scope of Barzel’s firm comprises the set of contracts whose variability is contractually guaranteed by common equity capital. The firm, then, is a nexus of outcome guarantees.

• 14. Sanjeev Sabhlok  |  10 February 2014 at 8:16 pm

  Sorry, haven’t read further comments, but thought I’d make note of this here for reference (I quote from Lee Ann Fennell’s The Problem of Resource Access (Harvard Law Review, Vol. 126))

  While agreeing with Coase that a zero transaction cost world would produce allocative efficiency, Demsetz views it as deeply mistaken to equate positive transaction costs, or rational reactions to them, with inefficiency. In a representative passage, Demsetz analogizes transaction costs to transportation costs:

  “Imagine a railroad capable of shipping goods between two firms. The railroad incurs cost if it does this, and the cost may be so high that the shipment does not occur (and, instead, as Coase wrote in ‘The Nature of the Firm’ (1937), the would-be receiving firm chooses to rely on in-house production of the good that would have been shipped were there no transport cost). No inefficiency has been created if the shipment does not take place under these circumstances, for the implied gain from making the shipment is less than the cost of doing so. But, pray tell, we reach the same conclusion if we change ‘shipment cost’ to transaction cost. So, we had better re-examine Coase’s reasoning about positive transaction cost.”

  In my view Demsetz is right. There are NO transaction costs. Only costs. We may define certain types (friction costs) as transaction costs but this is not good science. Economic science can only talk about costs (which include all types of costs), and I see the main implication of the Coase Theorem as the necessity to allocate property rights, not any need to worry about ‘transaction costs’. Whatever transactions are feasible and profitable, once property rights are allocated, taking account of costs and benefits, will occur, and will be allocatively efficient.

• 15. Jim Rose  |  10 February 2014 at 10:08 pm

  demsetz is an asute critic of transaction costs.

  My problem with transaction costs theories of the firm etc is the focus is stopping people ripping you off rather than something more positive such as implementing entrepreneurial judgement

• 16. Defining Transaction Costs | Economic Thought  |  19 February 2014 at 10:01 am

  […] Klein points out that the term “transaction costs” doesn’t have a definitive definition, and […]

• 17. Oscar  |  27 March 2014 at 6:54 pm

  Let me boil it down to the real world:
  Transaction costs are:
  Brokers
  Lawyers
  Lobbyists
  Assorted graft, bribery, and that ilk.

  Transaction costs are higher in the developing world than he developed world due to the last item.

• 18. Doug Allen  |  31 December 2014 at 6:37 pm

  Given that Peter was kind enough to point out my 2000 paper on Transaction Costs, that several of the comments have mentioned Yoram and Harold, and that a number of comments have been made that confirm Peter’s initial comment, I thought I might pipe in with a little personal history of thought on the topic and an appeal for the usefulness of my definition. [I apologize if this comes across as vain, but this might be the one forum where the readers actually care.]

  My first exposure to the question “what are transaction costs?” came on a 1982 midterm in Chris Hall’s advanced price theory class. [Chris Hall was a student of Cheung’s in the 1970s, and probably the most brilliant economist to come out of UW during the
  Cheung glory years. Alas he was never much of a publisher, but see his 1986 Econ. Inquiry paper on horse racing for a great paper to use in class to convince students of the power of economic thinking.]

  I wrote what I thought was a great answer and was surprised to get a 0 from Chris. When I asked him what the answer was he said “I don’t know, but I know your answer is wrong.” That started my quest for an answer … two years before I met Yoram in 1984.

  Yoram, of course, was working on the same issue, and took a crack at it in his 1985 JITE paper. That paper was also heavily influenced by Chris, but stops short of defining TCs. It does, however, establish the relationship between information costs and TCs. TCs are not “frictions”, they are a type of information cost.

  From 1986-1988 Yoram was working on his PR book, and I was working on my thesis. Chapter 2 of that thesis was called “What are Transaction Costs Anyway” the same title as this blog! I was also reading every word of Yoram’s book and making detailed comments. He was doing the same with my thesis. In private discussions we’ve noted several times that these ideas were joint efforts. However, I do recall the moment when it struck me that transaction costs and property rights are two sides of the same coin.
  This is the key to understanding the Coase Theorem, Institutional economics, and just about every other question of organization.

  I decided to drop the “Anyway” from the title of my chapter and shopped the paper around as “What are transaction costs?” I’ve never had a paper rejected so often, so quickly, and with so many dismissive remarks. My friend Tom Borcherding, then editor of EI wrote me back saying “you should be ashamed of yourself, sending me this paper.” [Years later Tom would tell me that his biggest mistake as editor of EI was rejecting that paper.]

  Finally I found a champion in Dick Zerbe. Unfortunately, almost no one has ever heard of his journal “Research In Law and Economics” where the paper was published in 1991. I was lucky to get the chance to write the 2000 Ency. article where I was able to expand and elaborate.

  Recognizing that TC are the costs of establishing and maintaining PRs has many benefits. First, it is a well defined definition and is operational (I’ve spent my entire career putting it into practice). Second, it is the one definition that is sufficient for the Coase T. When these costs are zero, PR don’t matter, when they are positive they do. Every critique of Coase (to my knowledge) fails because of an inadequate definition of TCs. Third, it helps to distinguish between “iceberg”, “frictional”, or “neoclassical TCs” that show up in so many models. Every mention of TCs that is nothing more than a “T” in a cost function, are not real TCs. [Here’s where Harold is wrong. When he talks about TCs he often has little more than a friction in mind. Ironically, when he writes about property rights, he’s almost always on the right track.] Finally, the definition helps to organize many of the other specific types of costs that are TCs (eg. measurement, etc.]

  Well, that’s a long post, and probably too personal. I’ll stop there.

• 19. Bas Percival  |  11 March 2015 at 4:30 am

  Professor Allen, thank you so much for the above. Anecdotal history is the best way in giving tricky theories a human dimension and as such making them easier to understand.

• 20. Peter Klein on Transaction Costs  |  9 July 2019 at 9:23 am

  […] Also, listeners may enjoy the comments on this blog post. […]

Trackback this post  |  Subscribe to the comments via RSS Feed