The Complementary Relationship Between Institutional and Complexity Economics: The Example of Deep Mechanismic Explanations

Abstract:

Analyzing economic systems from an evolutionary-institutional or a complexity perspective are two complementary approaches to economic inquiry. I discuss three arguments in favor of this hypothesis: (i) eminent institutional economists have examined the economy as what today could be considered a complex system; (ii) complexity economists lack meta-theoretical foundations which could be provided by institutionalist theory; and (iii) institutional economists could benefit from using methods of complexity economics. In this context, I argue that scholars considering the economy to be complex should seek to explain it by discovering social mechanisms instead of focusing on prediction. In order to distinguish between alternative explanations, scholars should refer to the deepness of an explanation, rather than to Occam’s razor.

Keywords::

• agent-based computational economics
• complexity economics
• evolutionary-institutional economics
• philosophy of science
• systemism

JEL Classification Codes::

• B25
• B41
• B52

Notes

¹ Mario Bunge (1999, 2012) provides a more extensive list of institutionalist economists who (implicitly) have worked in a systemist way.

² This might sound surprising given the poor predictive performance of many mainstream models. This lack of predictive capacity is usually “explained” by the violation of particular ceteris paribus assumptions.

³ “Instrumentalism” refers to the particular positivist epistemology. It does not refer to the institutionalist approach to theorizing and policy-making, as outlined in F. Gregory Hayden (2006, ch.3).

⁴ This is particularly the case if two explanations come from different, potentially incommensurable paradigms (Kuhn 1962). As economics should currently be considered a multi-paradigmatic science (see Elsner 1986), this is a probable incidence.

⁵ Since both models essentially give the same predictions on sustainable growth rates, the standard model is certainly simpler than the one offered by Witt. Defenders of Occam’s razor could then argue that Witt’s theory explains the observed facts better than the standard theory, and thus should be preferred. But then they would lack a concept explaining what they mean by “better.”

⁶ This process should be embedded into an adequate epistemological framework to clarify how the explanatory power of the model increases. I have previously suggested such a framework (Gräbner 2017), but there are other useful alternatives. The most important thing is that the epistemological framework is made explicit in the modeling exercise.

⁷ Alessio Moneta and Federica Russo (2014) discuss the concept of evidential pluralism that specifies the circumstances under which one can make informed statements about causal mechanisms in reality. They argue that both background knowledge on the mechanisms at work and statistical evidence is needed. Computational models are useful in supplying the mechanismic information in a form that can then be tested statistically.

⁸ Geoffrey M. Hodgson (2015, 139) makes a good proposal by defining a market as an organized and institutionalized recurrent exchange.

Additional information

Notes on contributors

Claudius Gräbner

Claudius Gräbner is a research associate in the Institute for the Comprehensive Analysis of the Economy (ICAE) at the Johannes Kepler University Linz (Austria).