The Diagnostic Triple


Sep 11, 2017 — Neil Boyle

The Diagnostic Triple came out of reading the January 2017 conference papers and proceedings of the American Economic Association’s the Richard T. Ely Lecture titled, “The Economist as Plumber”, by Esther Duflo.  Duflo was not writing her lecture with transaction cost economics (TCE) necessarily in mind although it supported some of the same themes.  I thought her concept of detail was close enough so that I expanded the idea to a triple by adding “process” and “empirical experimentation” as the second and third components and called the initiative, “The Diagnostic Triple.”  I have not tested the diagnostic triple in any rigorous manner, e.g., in terms of finding an empirically adequate sample in real life, or in terms of its empirical relevancy or robustness.

One of my interests in governance is the search for mechanisms (or institutions) that can transform what Williamson refers to as theory into instrumentalities of efficient economic organization.  An efficient economic organization is what TCE is about.  To this extent, I adapted Duflo’s concept of detail into the “diagnostic triple” comprised of the following three components: detail (insufficient and abstracted information are constant inputs of traditional economic modelling, for example, most data are sourced from the polity rather than from the transaction); process (is almost completely ignored in traditional economic analysis and when process is included it is largely sourced from the economics of price theory, which often leads to maladaptation); and empirical experimentation (is inadequate and rarely done and what is done is dominated by price theoretic considerations hence resulting in price theoretic patois and keeping the organization of the phenomena one is attempting to model at bay and distant so that learning is impaired).

Before proceeding further with explaining the diagnostic triple, I want to expand a bit on the concept of theory because it relates to the topic of understanding TCE and some of the “bad habits” we have gotten into.  The practice of avoiding theory has been around for some time.  I know because I have experienced this form of avoidance throughout my 50 year career in the field of international development.  It is as if theory-translation into practice-mechanisms are the province of only a few highly credentialed individuals.  Perhaps this is the case because most professionals are not aware of how to translate.  Witness the following two paragraphs: John Briscoe’s comment during a conference I attended about the use of theory and the performance of World Bank practitioners and researchers that “practitioners practice but don’t read, and researchers read but don’t practice”. (Briscoe. 2015)  (John Briscoe (now deceased) was the head of the Water Group in the center at the World Bank when I was still at the Bank.  The Group was responsible for setting water policy for the Bank.  A South African, John was honored by the Nobel Committee in Stockholm for his work in water, a precious worldwide resource.)

When reading Ronald Coase’s remark (that follows subsequently), please read the descriptors as identical to Briscoe’s comment above, that is, Briscoe’s researcher should be read as identical to Coase’s theorist and Briscoe’s practitioner as identical to Coase’s institutionalist.  Ronald Coase remarked, “[W]e have less to fear from institutionalists who are not theorists than from theorists who are not institutionalists.”  This was in 1964, page 196.  Thirty two years later Williamson replied, “It is both possible and desirable to combine institutional economics with theory, and the time has come to do precisely that.” (Williamson, 1996:342) The large elapse of time of 32 years between Coase and Williamson I read as an indication of the evolution of knowledge that relevant knowledge was not available during those 32 years.

The diagnostic triple.  I assert that the application of the diagnostic triple to a typical infrastructure project reveals a perspective of how and why economic agents organize as they do, which is not anything similar to the way economists or researchers (theorists) model their economics or the way institutionalists (practitioners) analyze institutions.  The how and why are two important questions for anyone wanting to plumb what “organization and governance” are about.   Process incentives are brought to light, human behavioral based-relations between government buyers and private suppliers and operators (and others) are made scientifically operational, and the contractual agreement and structure between the two principals takes shape and comes into the light through experimentation.  These, of course, are just three of the many revelations that accrue to the diligent practitioner; there are more.

The triple originates from the application of TCE to my work over the past 50 years, but this begs the question, Do I know TCE?  I spent more that 12 years studying and writing about theory and 50 years applying TCE and for six of those twelve years Nobel Laureate Oliver Williamson was an advisor to me; it took six years before Williamson was sufficiently comfortable to endorse my work.  For those who are interested in pursuing my TCE competence further may peruse my website at for my CV and samples of papers I wrote under his supervision.  Some of these papers were reviewed by Sharon Poczter a former top student of Williamson’s who is now assistant professor of economics at Cornell University.  The components of the triple are described as follows:

Detail – Economist are asked more and more to do more than analyze policy prescriptions, but also to design the institutions so they work.  This entails a mindset that the usual economist is not particularly trained for.  The economist (without more) struggles to get to the level of detail of a plumber or in a different way, an engineer.  Plumbers install machines, watches what happens when they operate, and then tinkers as needed until the machine functions optimally.  This puts the plumber in close contact with the mechanisms involved and with the phenomena in general. By phenomena, I refer to the man-machine system and how the two empirically work together, dynamically.  For instance, it is not just the rate and volume of flow of water through a pump that boosts the flow through a conduit, it is also the placement of the conduit in the network of conduits, the material composition of the conduit and its flexibility, and both of these are set by humans and are subject to human error and misalignments with other parts of the institutional and production systems.  The plumber gets to know the phenomena better than the scientist and the engineer, and certainly better than the economist, who models the phenomena rather than engages it as the plumber does.  By contrast, the engineer takes the general principles of science and applies them to a specific problem and pays particular attention to the specifics of the environment.  She also pays attention to new tools if needed, for instance, in the case where she cannot ignore the fact that a “particular situation is not covered by the assumptions of the theorem, and cannot ask agents to change their preferences so the assumptions hold.” If the problem cannot be solved analytically, the engineer will try other tools—typically “computational tools and laboratory experiments—and will simulate the behavior of a market.”  (Duflo, 2017:____)  I believe this is a step in the right direction but still a far cry from engagement and alignment.

Process – This is the “how” one gets to achieve his objective; it is not the objective per se, nor is it the “what price is it” kind of question.  The “how” was the question posed by Adam Smith during his study of the pin industry in 18th century Scotland.  Smith was interested in how the factories were organized.  As a result he discovered several organizational aspects that are crucial in today’s world: division of labor, specialization, economies of scale, assembly line production, and self-interest as opposed to the dictates of the monarch.  He also determined that the price of a pin depended on the way the factory was organized.  Williamson (2010)

Unfortunately Smith ignored the human element in the production process.  He did not possess the knowledge that was necessary to do so; it hadn’t been discovered yet.  Smith recorded what he had learned about the pin industry, but overtime they were interpreted by economists in “production language”, a form of economic organization patois—roughly 114 years later when Alfred Marshall published his “Principles of Economics” in 1890.

Empirical experimentation – I consider this third component of the  triple the most important because it keeps the phenomena constantly in the sights of the practitioner studying it; it keeps the phenomena close at hand so that the practitioner-economist and the phenomena can interact and learn from each other.  How?  The answer is by what I call “next-step” experimentation. Next-step experimentation is institutional change that takes place within the next project. It is where feedback from the contemporary project is fed back into the next project.  It is what clinical psychologists call “action research”.  (I was introduced to “action research” when I worked at the NTL Institute for Applied Behavioral Science. This was my first formalized training in institutional change.  NTL had the elements of change that I was searching for: theory-based change, e.g., Kurt Lewin’s Force Field Analysis for social change within large communities of people–large social groupings.  NTL initiated the field of organizational behavior, a field offered by most MBA programs in universities around the world.  In the late 1940s and early 1950s, NTL invented sensitivity training as a solution to racial conflict between blacks and whites in Hartford, Connecticut.)  Sensitivity training is oriented toward the personal growth of individuals and small groups, organizational behavior and organization development (OD) is oriented toward the growth of the firm and institutions, and social change is oriented toward larger communities of people than firms or institutions such as Hartford, Connecticut.

Until the problem of renegotiation in the contemporary project is resolved with the absence of moral hazard, experimentation will be limited to the next project.  Let me explain.  The necessary alignment can be done in the current project but only if renegotiation can be made viable.  Research shows that renegotiation is a fundamental mechanism to achieve ex-post realignment, and ex-post realignment assures performance (ceteris paribus) whenever the agent “falls off the contact curve”, which occurs routinely.  The problem is renegotiation creates moral hazard at the time of bidding because bidders lowball their bids knowing they can recover their lost revenues at renegotiation.

Empirical experimentation generates refutable hypotheses (testable predictions) that can be tested by economists.  It is a form of tinkering (like the plumber in Duflo’s lecture) to correct errors that brings the phenomena into alignment.  In engineering terms, this kind of controlled experimentation works because in electrical engineering the double feedback loop does the job of “renegotiation” except they do it in the contemporaneous project.  The double loop is two feedback loops, one for performance, another for negative error sensing, which assures performance. Double feedback loops are akin to controlled experimentation.

Diagnosis begins with the generation of a refutable hypothesis that the triple is in disequilibrium and the contract is unstable as the cause of disequilibrium.  With an unstable contract and a triple that is in disequilibrium, the solution is to bring the components of the triple into equilibrium. This is accomplished in one of two ways: 1) renegotiations during the ex-post implementation interval; or 2) by following a process of conflict resolution in-its-entirety where conflict is generalized to include any problem that arises between parties that holds-up the continuation of work. Renegotiation without the moral hazard of lowball bidding is possible only if renegotiation occurs not in-its-entirety but only during the ex-post implementation interval, and then, only if price changes are permitted through intense verification of asset specificity (k) and safeguards (s).  For the contemporary practitioner, however, this framework is counterintuitive. It is not readily apparent without preparation.

Tags: , , , , , , , , ,

This article is filed in articles. Start or participate in a discussion below or click here to see more articles in this category.


Please share your thoughts and comments about the article. We look forward to hearing from you. Fields marked with "*" are required: