Evolutionary biologist Leslie Orgel’s second rule of biology states that “evolution is smarter than you are.” Evolution and its mechanisms can explain much of the seemingly boundless complexity and organization evident in biological ecosystems. As I discussed in my last essay, biological evolution with its objective function of survival and reproduction driven by the processes of natural selection operating on variation caused by random events (mutations) and sexual reproduction moves populations towards locally adaptive peaks on a fitness landscape. Therefore, might there be lessons that can be extrapolated from evolution that can be applied to the “design” of social systems? In our seemingly precision-engineered world of causes and effects, inputs and outputs, it might be counterintuitive to consider that in social systems outcomes are often not or at best weakly correlated to intentions. Historically, it is far from obvious that humans can intentionally design effective long-standing institutions and organizations. Empirically, it is not clear whether good intentions with an illusion of precision, or evil intentions, have caused more harm in the world. In a constantly changing environment with multiple conflicting and incommensurable objectives, where full information is never available and the problem is never fully defined, an evolutionary perspective of variation, adaptation, and differential reproduction may lead to better results than narrowly defined objectives and metrics.
The lessons of biological evolution and parallels to social system “design” can be foundational and far-reaching. In biological evolution, population genotypes and phenotypes are the substrate upon which natural selection operates upon. The fittest – measured by the objective function of survival and reproduction – are selected and the least fit is removed from the population. Furthermore, the variation is the rule and itself selected for by natural selection. This diversity in substrates is critical not only for discovering higher adaptive peaks but also for providing insurance in shifting environments. Without this diversity, evolution is limited in its range of solutions and thus has limited robustness against shocks. Analogously for social systems, the raw material is ideas and hypotheses and the selection mechanism (in our modern world) is the market or the majority. The fittest ideas are selected and the less fit ideas are removed from the population. Similar to biological systems, social systems that encourage and foster thought diversity are resilient against shocks and are more likely to move to higher peaks on a shifting adaptive landscape. Emulating evolution, the processes of small scale experimentation, iteration, discovery, and adaptation are probably the best means at our disposal for building adaptable systems in a complex, shifting, and uncertain environment.
However, in order to experiment and iterate there has to be a criterion to test against and a measure to achieve. The criteria for biological evolution is the relatively simple and straightforward mandate to survive to reproduce. Much of the biological complexity and organization can be explained by evolutionary processes optimized to this criterion. However, because of the problem of the criterion, the reduction of the objectives of social institutions to narrowly defined criteria is maladaptive, oversimplifications, and misleading. In contrast to evolution, social system-wide objectives must be loosely defined as there is not a single best way of achieving them and there is not a single measure of defining them. For example, I would argue, the healthcare system should have a high-level objective of improving health. However, what does it mean to improve health? How do you measure improved health? Is it life expectancy, disability affected life years (DALY) or quality-adjusted life years (QALY)? At what time scale is “improving health” pertinent? How do you balance antibiotic stewardship at the health system level with treating an individual patient with an ambiguous infection that is most likely viral in etiology? How do you consider the societal impact of the opioid epidemic with the pain of an individual patient at the bedside? There is no single answer to these questions and thus the best option we have is to set high-level objectives and achieve them asymptotically through adaptation and iteration, with the constant rebalancing of seemingly incompatible and incommensurable objectives.
The healthcare system — like all complex adaptive systems — is a highly dynamic system that is a product of its subcomponents but cannot be defined by or reduced to its subcomponents. As is often said, the “whole is more than the sum of its parts.” A core lesson that can be gleaned from history is that although the future of social systems is a product of the past, they are not necessarily the vision of the past. They are often what the past unexpectedly and unpredictably enabled. Since the future is uncertain and imperfectly understood, the process of experimentation and iteration is usually more robust and better adaptable than a rigid, narrowly defined objective-driven system. As the 18th-century Scottish philosopher, Adam Ferguson wrote, “nations stumble upon establishments, which indeed are the result of human action, but not the execution of human design.” High-level loosely defined objectives, the plurality of ideas, competition, coexistence, experimentation, coevolution, and adaptation are all concepts that can be extrapolated and applied from evolutionary theory to help us “design” social systems.