Partial equilibrium and the advent of non-mathematical quantitative methods
Marshall’s partial equilibrium approach remained in the shadow for long after Sraffa’s 1925 critique, but did not entirely disappear. The static supply-and-demand diagram was used in some applied studies, while the dynamic conception of sequential transactions between pairs of buyers and sellers, underdeveloped in Marshall’s own time, was refined with the help of novel, non-mathematical techniques.
It underpinned the pioneering laboratory experiments on markets that were implemented by Edward Chamberlin (1948) and Vernon L. Smith (1962). Although experimental research remained a tiny niche within economics for a long time, it kept alive the Marshallian approach in its dynamic connotation. Today, partial equilibrium experimental markets are still of great interest and the properties of the exchange dynamics that they generate are often referred to in the literature as “Marshallian dynamics” (Friedman and Rust 1993).Interestingly, the first experimental markets inspired by Marshall’s dynamic orientation were being devised at the same time in which set-theoretical mathematical techniques were being implemented to model Walrasian systems. These tools were rudimentary at the beginning (Chamberlin simply relied on his students and a blackboard to perform his market experiment), but they became more powerful over time with the increased availability of software for computer-based experiments and the possibility to perform parallel experiments with human and artificial (simulated) agents (Gode and Sunder 1993). They correspond to a view of economics that gives less weight to the logical rigour of formalization than to the theory’s fit with empirical data, assessed with objective, well- defined and replicable measures. While Walras’s approach and its mathematical bases experience a slow but neat decline, experimental and computational approaches are on the rise, transforming the discipline.