Saved in:
Bibliographic Details
Main Author: Smirnov, Roman G.
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2506.00032
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908387670753280
author Smirnov, Roman G.
author_facet Smirnov, Roman G.
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2506_00032
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Deriving Production Functions in Economics Through Data-Driven Dynamical Systems
Smirnov, Roman G.
Applications
91B55, 37N40
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.
title Deriving Production Functions in Economics Through Data-Driven Dynamical Systems
topic Applications
91B55, 37N40
url https://arxiv.org/abs/2506.00032