The Scientific Status of Economics: Models, Laws, and the Challenge of Reality

This podcast episode critically examines the scientific status of economics, using Plato's allegory of the cave to question whether economic methods truly allow economists to grasp reality or merely trap them in theoretical constructs. The discussion centers on how economics attempts to establish "laws" through modeling, which involves inputs, logical processes, and outputs. It explores three contrasting philosophical views on model building: starting with realistic assumptions, prioritizing logical consistency, or focusing solely on predictive success. Jevons' 19th-century advice to "start with the facts and end with the facts" is presented as a foundational, albeit often unfulfilled, ideal for model construction, involving inductive hypotheses, deductive conclusions, and empirical testing.

The episode delves into the significant challenges encountered at each stage of model building. Regarding assumptions, it contrasts Jevons' call for realism with Nicholas Kaldor's concept of "stylized facts," noting that many mainstream models often derive assumptions arbitrarily. For the logical stage, while mathematical rigor is crucial (as emphasized by Robert Lucas), the podcast argues that logic alone cannot describe the real world; it must be grounded in realistic hypotheses. The most profound problem lies in testing predictions against reality, particularly the difficulty of falsification in economics. Unlike natural sciences, economics cannot conduct controlled experiments on entire economies, and its primary substitute, econometrics, suffers from inherent weaknesses.

Further elaborating on empirical limitations, the discussion highlights the two major problems with econometrics identified by Tinbergen: the near impossibility of isolating a single hypothesis for testing and the issue of non-stationary time series. Simulation modeling is also critiqued, often yielding results that are largely predetermined by its premises, leading to a "garbage in, garbage out" scenario. While randomized controlled trials (RCTs) have proven effective for evaluating micro-level public policy interventions, such as Mexico's Progresa scheme, they are deemed inadequate for testing macro models. The episode then explores "post-modernistic economics," which views economics as a branch of literature or "persuasive utterance," where mathematics serves as a rhetorical device rather than a tool for objective proof, as argued by Philip Mirowski and Deirdre McCloskey.

In conclusion, the podcast asserts that economics does not function as a natural science because it cannot compel facts to conform to its theoretical assumptions. Its models frequently fail to account for observed phenomena, and despite increasing precision in its predictive apparatus, it has not improved its ability to make specific predictions, remaining largely confined to "generic" or qualitative forecasts (Rosenberg). The fundamental weakness stems from economic causal laws operating through human agency, which inherently lacks perfect knowledge of the future. Therefore, models that assume such perfect foresight are ultimately deemed "science fiction," and the persistent, often angry, debates among economists are likened to theological disputes, lacking definitive scientific resolution.

do the methods of economics allow economists to escape from the cave or do they imprison them in the shadows of their own theoretical reasoning

building a model is like drawing a map the object is to leave out cluttering information while leaving in place the crucial information

a model that is just as complicated as the world it is designed to portray isn't going to be any use

start with the facts and end with the facts

the theorist in my view should be free to start with a stylized view of the facts i.e concentrate on broad tendencies ignoring individual detail and proceed on the as if method

economics can't live by logic alone to be useful a logical theorem needs to be based on a realistic hypothesis logic can tell you nothing about the real world it can only tell you about itself

you can never verify an inductive hypothesis it can only be falsified and so science he suggests proceeds by successive falsification

it's impossible to experiment with whole economies whole populations and second because of the weakness of the substitute for experiment which is econometrics

garbage in garbage out is certainly appropriate

economics like other social sciences indeed like the hard sciences as well is built from persuasive utterance

if economic propositions were provable economists and others wouldn't need to pay attention to confidence is required for stories not for science

the weakness of economics as a hard science stems from the fact that its causal laws have to operate through human agency

models which assume perfect knowledge of future events are not science but science fiction