Why Transaction Cost Analysis is Attractive, and Very Flawed
This piece, "Markets Are Eating The World," is being shared widely. The author, Taylor Pearson, tells a story of the economic history of markets and firms, and what comes next, explained in terms of transaction costs. It's a fun read, with a seemingly insightful narrative, and makes several good points.
But while the transaction cost explanation is intuitive, it is also mostly wrong. It places the cart in front of the horse.
I fully understand the attractiveness of using transaction costs, especially in the colloquial (imprecise) sense, because it appears so powerful. Transaction costs explain so much.
For example, in this article, the author notes that "The lower the transaction costs are, the more efficient markets will be, and the smaller firms will be." What a beautifully simple explanation!
And, as it seeps in between the lines throughout the article, most of economic history can be explained using the same concept: transaction costs explanation seems to explain almost everything.
The problem is that something that appears to explain everything actually does not explain anything at all. Dressing up the history of the world, and its economic development, in terms of transaction costs is mostly not very helpful.
The reason is the flaw already existent in Coase's Nobel-winning 1937 article: the existence and size of the firm is explained using a simple calculus of costs transacting within the firm (organizing costs) and outside (transaction costs). But while this seems obvious, and in some sense it is, it assumes a certain degree of specialization/division of labor (Coase notes the "specialized economy") that is unchanging. Because if you engage in increased specialization, then there can be no calculus at the margin.
This applies to most examples in the article, for example, Ford's factory.
Did Henry Ford have the option of contracting with free agents instead of integrating production in his own factory? No, because his production process was innovative and therefore it could not have been accomplished by overcoming triangulation, transfer, and trust transaction costs. There was no such thing as assembly line manufacturing of automobiles, so it had to be created by Ford.
Some not very wise wiseguys might opine that this is explained by transaction costs for the factors necessary to establish market-based (through contracts) manufacturing of the Model T is practically "infinite." That's a terrible analysis since we're basically tweaking the concept to provide an explanation no matter what situation.
The fact is that Coase, in a sense, understood the limitations of his analysis.
This is why he assumed an already specialized market with rather standardized factors that could be organized by the price mechanism (market) or a manager (in the firm). If things are standardized, much like building your company using LEGOs, then it truly becomes a cost calculus at the margin.
But how often does this actually apply to entrepreneurs? Very rarely. They apply to some degree to firms in commodity markets, but less the more specialized (differentiated) firms are.
For entrepreneurs and innovators, Coasean transaction cost analysis explains almost nothing, because it is not about choosing a least-cost means for coordinating factors in the already decided production process. It is about creating it, at least partially de novo.
As Coase also, albeit implicitly, realizes, the transaction cost explanation requires not only that actors do not specialize, but also that specialization has already happened. This is effectively a static state analysis. Granted, one can shoe-horn Coase's conceptual apparatus into other situations, but that's fair neither to Coase nor very helpful.
The fact is, what Pearson is explaining is a process, a flow of continuous (but not steady) change — and the tool he uses (transaction costs) is for situations where nothing (except transaction costs themselves) change. The result is the appearance of insight, but in reality he's only adding seemingly clarifying terminology while creating confusion.
While the overall process is described well, and the story is compellingly written, the use of transaction costs largely misrepresents what happened. Transaction costs seem to explain the situations after the fact, but don't explain the actual goings-on or the decisions made. The framework looks at the effect and presents it as the cause.
This is, of course, more problematic than academic quibbling about definitions, because it may make entrepreneurs think their enterprise will be successful if they can find the lowest transaction costs, whereas the truth is that lower transaction costs are a possible outcome of their actions, especially after the market settles (for example, after a disruptive innovation has been fully adopted by the market).
Of course, transactions costs affect how you run your business, as do any costs, but they're not the reason you have a firm or predict the better way of organizing it. You need different tools for this, and you need a process perspective.
I elaborate on this in my book The Problem of Production, which started out as a critique of the over-reliance and over-application of the transaction cost concept, but turned into a competing theory of the firm and economic development — one that acknowledges process, change, and entrepreneurship. And that doesn't use the horse to push the cart forward.