Deriving Production Functions in Economics Through Data-Driven Dynamical Systems

TL;DR

This paper introduces a dynamical systems approach to derive production functions from economic data, revealing how classical forms like Cobb-Douglas and CES naturally emerge from growth dynamics.

Contribution

It presents a novel method that models economic growth as dynamical systems to systematically derive and verify production functions from data.

Findings

Cobb-Douglas form emerges from exponential growth dynamics

CES production function is a special case of fundamental invariants

Method bridges statistical analysis with systems theory

Abstract

In their seminal 1928 work, Charles Cobb and Paul Douglas empirically validated the Cobb-Douglas production function through statistical analysis of U.S. economic data from 1899 to 1923. While this established the function's theoretical foundation for growth models like Solow-Swan and its extensions, it simultaneously revealed a fundamental limitation: their methodology could not determine whether alternative production functions might equally explain the observed data. This paper presents a novel dynamical systems approach to production function estimation. By modeling economic growth trajectories as dynamical systems, we derive production functions as time-independent invariants -- a method that systematically generates all possible functional forms compatible with observed data. Applying this framework to Cobb and Douglas's original dataset yields two key results: First, we demonstrate that the Cobb-Douglas form emerges naturally from exponential growth dynamics in labor, capital, and production. Second, we show how combining fundamental invariants of this exponential system generates the CES production function as a special case. Our methodology bridges statistical analysis with mathematical systems theory, providing both a verification mechanism for classical results and a tool for discovering new functional relationships.