Cartels of Mediocrity

Reproducing a sociology simulation paper about social norms and the tendency towards low-quality exchanges and suboptimal outcomes.

The Unaccountability Machine1 by Dan Davies is broadly about how and why dysfunctional systems produce outcomes nobody seems to want.<br>It also contains a tidy introduction to Cybernetics2 and ideas like the viable system model3 and requisite variety4.

Reading that got me interested in more systems and sociology topics, and eventually led me to these papers:

The LL game: The curious preference for low quality and its norms (Gambetta & Origgi, 2012)5

Social Norms and the Dominance of Low-Doers (Proietti & Franco, 2018)6

Despite knowing nothing about game theory, sociology, or behavioral modeling, I thought both papers were approachable and engaging.<br>After 25 years in industry it was refreshing to see a formal academic take on organizational cliques.<br>And the papers were timely with a few things in my life:

I’d been prototyping some game ideas involving simulated agent populations and emergent behavior with my kid, and the second paper does exactly that

I’d been looking for a reason to try Clerk7, a Clojure visual notebook tool

It was performance review time at work

So I committed to reproducing the paper’s findings in Clojure and its tables/figures (plus new ones) using Clerk:

Clojure source: https://github.com/taylorwood/low-doers

Clerk notebooks: fundamentals and figures

The phenomenon

Professionally, have you ever felt like your hard work wasn’t furthering your career, or even irritating some colleagues?<br>Ever thought things might be easier if you coasted a bit?<br>Maybe you needn’t even feel bad about it if your peers had the same mindset.

The dissonance is reduced by interacting always with the same people, whom one can trust for not challenging one’s standards. L-doers segregate themselves in mutual admiration societies.

This is the mindset from which Gambetta & Origgi’s “cartels of mediocrity” arise. The LL game abstract:

We investigate a phenomenon which we have experienced as common when dealing with an assortment of Italian public and private institutions: people promise to exchange high quality goods and services (H), but then something goes wrong and the quality delivered is lower than promised (L) . While this is perceived as ‘cheating’ by outsiders, insiders seem not only to adapt but to rely on this outcome. They do not resent low quality exchanges, in fact they seem to resent high quality ones, and are inclined to ostracise and avoid dealing with agents who deliver high quality . This equilibrium violates the standard preference ranking associated to the prisoner’s dilemma and similar games, whereby self-interested rational agents prefer to dish out low quality in exchange for high quality. While equally ‘lazy’, agents in our L-worlds are nonetheless oddly ‘pro-social’: to the advantage of maximizing their raw self-interest, they prefer to receive low quality provided that they too can in exchange deliver low quality without embarrassment . They develop a set of oblique social norms to sustain their preferred equilibrium when threatened by intrusions of high quality. We argue that cooperation is not always for the better: high quality collective outcomes are not only endangered by self-interested individual defectors, but by ‘cartels’ of mutually satisfied mediocrities .

And later in the LL game:

Our basic point so far can be summed up thus: if you give me L but in return you tolerate my L we collude on L-ness, we become friends in L-ness , just like friends we tolerate each other’s weaknesses. But if you give me H that leaves you free to disclose my L-ness and complain about it . So you are not my friend, I fear and resent you, and if I cannot punish you for producing H, at least I avoid dealing with you. While in an ordinary world it is L-doers who are punished by avoidance and exclusion, in an L-dominated world it is H-doers who are ostracised. Essentially, the L-exchange can be seen as a cartel of mediocrities who pretend to be H .

There seems to be two forces that could contrast our supposed natural inclinations to L-ness and promote quality, one is the passion for a job well-done, the intrinsic pleasure found in employing and testing one’s skills at some task; the other is competition, succeeding at which carries extrinsic rewards . Generically, these forces fail if the algebraic sum of rewards and punishments for H-ness is lower than the sum of rewards and punishments for L-ness. Even H-prone individuals are ultimately driven to choose L (or to become eccentric and isolated ‘perfectionists’ or to migrate) if they systematically fail to gain any reward from their effort . In short, L spreads if it pays off. This remains a tautology, however, unless we can understand the conditions that affect the relative payoffs of H and L.

I recommend reading the paper just for the examples of different H/L exchanges, their...