Governance mechanisms and economic engines

In which we separate economic efficiency from integrity and accountability.

The focus of EconPatterns is on economic design, which inevitably involves but is not the same as economic modeling.

Design is in a way a precursor to modeling, in that it shapes how the model is conceived, to which extent it can provide an answer which question, and to which extent the answer can bear the full load of the question, or if it can only give a conditional answer under tight (but often ignored) limitations.

In other words, and to apply a pattern from the first newsletter, the role of design is to structure the question in a way that we can route the inevitably time- and resource-intensive effort of formal modeling towards its best use.

Everyday economic interaction in industrialized countries will almost inevitably involve some form of interaction with computers, even if we think of seemingly “fully analog” examples like paying in cash for organically grown produce at the farmers market. There’s a good chance that the cash came out of an ATM, and there’s an even better chance that the farmer used computers to plan the crop cycle and account for the proceeds.

This pretty much means that economic design almost inevitably involves formalized expressions, mathematical, statistical, computational, in order to build the economic edifices tackled to guide the interaction. This might be the default mode and the default target demographic of this newsletter.

But — and this is one of the key takeaways of this newsletter, so it bears repeating — the computational abstractions hold a striking resemblance to real physical structures serving a similar purpose, and whether it's an agora, a bazaar, a souk, a coffee house, a bodega, a farmers market, a modern supermarket, or an e-commerce website, the underlying structural components are remarkably similar.

Where EconPatterns differs from other approaches to economic design is that it tries to express these abstractions in design, architectural, or even engineering terms.

One key example is the distinction between two abstractions, introduced here as “governance mechanisms” and “economic engines”. Both are of the same type, so we might as well speak of “governance engines” or alternatively of “economic mechanisms”, but their roles are fundamentally different, so it makes sense to keep the nomenclature as far apart as possible.

Economic engines

An economic engine is any physical or abstract edifice that serves a defined economic goal.

A market is an economic engine, but also a recommender system, a sales outreach process, a trading bot, a production schedule, an auction, a staffing plan. Colloquially speaking, any engineered edifice where the ultimate objective function can be expressed in financial terms.

Economic engines serve one principal and are engineered to that end: to achieve economic efficiency.

Economic efficiency is related to, but distinct from physical or engineering efficiency. In a shorthand we could say it translates these types of efficiency into economic efficiency, but this point will get closer attention later.

Governance mechanisms

If we think of economic engines as a summary term for devices that provide economic benefits to their principals, we can contrast them against governance mechanisms as devices that benefit a larger group of participants, including multiple principals with diverging interests.

This connects us to the difference between decision theory and game theory, which in its early days was often referred to “multi-principal decision theory”.

Especially if we think of economic engines as mechanisms under unilateral control, governance mechanisms are mechanisms under mutual control. This is an attractive proposition but as it turns out, exceedingly hard to implement in practice.

To take exclusive control of a profitable economic activity with multiple participants, and to extract rewards beyond one’s contributions (“rents” in econspeak) is a very tempting thing, so designing systems of mutual control — checks and balances in politics — to counteract this individual impulse, is a design problem of a higher order than designing economic engines.

There is a perennial risk that governance mechanisms, designed to benefit a larger group of participants, descend into mere economic engines, jerry-rigged to benefit a smaller group of well-positioned and opportunistic group of participants.

But beyond looking at a given economic activity from either one or from multiple perspectives, there is another, rather less investigated reason to introduce another type of mechanism.

And that is, if the goal of an economic engine is to achieve efficiency, the goal of a governance mechanism is to achieve integrity.

Efficiency vs integrity

“Integrity” is one of those positively connotated terms that are always at risk of being coopted by the buzzword merchants, so it might be worth giving a more precise definition, which helps us in this endeavor.

If we think of efficiency as “best use of the given resources”, integrity is the combination of efficiency and accountability. It is the state where all contributions have been delivered, or in other words, all expectations have been met, and all conflicts have been resolved.

Accountability is the state where we can map all contributions, positive or negative, to their contributors. In other words, when integrity breaks down, accountability helps us find out why and where it broke down.

This brings us back to our patterns from the first newsletters, namely that most economic activity is a network of efforts concatenated across multiple participants, within or across organizations, and economic design is inevitably concerned with orchestrating these efforts towards a good end.

Making good use of one’s resources is clearly such a “good end”. But so is returning rewards towards contributors according to the positive impact of their contributors on the ultimate work product. This is a much trickier question than at first thought, and it is not sufficiently captured with the efficiency criterion alone.

This will clearly become a focal point of EconPatterns, but it turns out that introducing the concepts of integrity and accountability gives us a very different lever to address questions of economic organization, especially given the short shrift they’ve so far received from economic theory, we can put them to good use and turn them into design patterns and ultimately better design.

This becomes even more pressing as states are increasingly struggling with the social impact of new technologies, and are very counterproductively trying to quell negative consequences with increasingly cumbersome and convoluted regulatory intervention mostly designed around burdening an already overstretched legal system.

If economic engines are becoming increasingly automated and are increasingly able to conduct (and transform) economic activity at scale, we have to come up with ways to design governance mechanisms that can keep up with this development and that curtail rather than attempt to penalize self-interested activity by privileged actors.