Why realism and methodological pluralism matter for robust research and public policy: perspectives from behavioural economics

Abstract

Conventional economics maintains that a critical test of the veracity of robust economic theory is its capacity to generate plausible economic predictions, irrespective of the realism of the theories' underlying assumptions. This methodological argument even holds for relatively less conventional approaches to economics such as behavioural, heterodox, experimental, and institutional. Following upon research in behavioural economics, I argue that such a methodology can easily result in the illusion of causality, the omission of potentially key variables, and closing the doors to key analytical questions as well as to publication bias. This generate perverse analytical results, with severe consequence for public policy. I argue that methodological pluralism is critical to the construction of robust economic theory, irrespective of ones' political orientation. Examples are drawn from financial markets, labour markets, and macroeconomics to illustrate this pluralistic perspective to economic analyses.