The following was written together with colleagues from ASU’s Biosocial Complexity Initiative and the IASS Potsdam. It was inspired by my previous post on complexity economics, which initiated a brief Twitter exchange with an economist from INET Oxford. Furthermore, it is addressed to the Forum for New Economy in Berlin, especially Michael Jacobs’ plea for a new paradigm in economic thinking.
In the spirit of the new economic paradigm put forth by Michael Jacobs at the Forum for a New Economy Launch this past Halloween, we would like to offer some insights from complexity studies for thinking about this new paradigm.We hold that the economy is a quintessential example of a complex adaptive system (CAS), characterized by multi-level interactions between learning agents operating with incomplete information, emergent patterns, and path dependent development. The overall economic system can be conceived as multiple systems which partially overlap one another, encompassing what might be traditionally delineated as social, political and environmental realms.
What can this conceptualization of the economic system tell us? Here we offer three specific examples of how complexity studies could contribute to a new economic paradigm. The first is a pragmatic recognition of the limitations of predictions within complex systems. Friedrich von Hayek, in his 1972 Nobel Speech, The Pretense of Knowledge, admonishes economists for their use of scientific methods developed for studying relatively non-complex physical systems to analyze complex economic systems, and the profession’s fixation on achieving precise numerical solutions to economic problems rather than trying to understand the underlying way the economy works, albeit in a descriptive and less precise way. “I prefer true but imperfect knowledge, even if it leaves much undetermined and unpredictable, to a pretense of exact knowledge that is likely to be false.” In this regard, we believe that CAS reinforces skepticism in our ability to predict the outcomes of economic policy and the confidence with which the economics profession often describes both the world and itself.
Another contribution from complexity studies is the concept of emergence, whereby the sum is greater than the parts and micro interactions within a system lead to macro outcomes that are qualitatively different from the behavior of the individual agents. This sheds doubt on the micro-foundation framing of traditional macroeconomic modeling, such as rational, representative agents maximizing utility in a general equilibrium setting, and suggests that analyzing the individual behavior of economic agents or sub-systems may not be a feasible way for understanding the larger system. From a complexity standpoint, a more promising method for building macro effects out of micro interactions are via agent-based models (ABMs), in which we see how heterogenous agents acting at the micro level can lead to novel macro outcomes.
Finally, a CAS is evolutionary and path dependent, thus undermining the relevance of econometric work focusing on identifying key variables in past economic trends. Changes in an economic system consist of the interplay of human agency within the context of changing regulation, innovation, societal values, and vested interests in a constantly evolving system. As new states in the system are obtained, the state space transforms, opens up new space for the system development, and closes off previously possible state spaces. As a result, the system continues its open-ended development in novel and unpredictable ways.
We do not intend to say there is no value in econometrics or DSGE equations. Rather, we believe a new paradigm will be best facilitated through a plurality of approaches and perspectives, in particular complexity thinking and the epistemological and methodological lessons it brings. This plurality is essential both within the broader context of economic thought and within the nascent field of complexity economics itself. We observe that one prominent strand of complexity economics situates itself somewhere between the traditional neoclassical paradigm and a claim to being “non-theoretical and data driven,” employing complexity modeling techniques such as ABMs, networks, and computational game theory. While we see this as a positive development, current work still shies away from applying the full set of complexity studies thinking to economics issues, such as providing co-evolutionary explanations for economic change and including vital economic elements such as money. Most importantly, all economic thinking needs the epistemological humbleness that a solid understanding of complexity should bestow.