Imagination Rather Than Observation in Econometrics: Ragnar Frisch’s Hypothetical Experiments as Thought Experiments

Hopos: The Journal of the International Society for the History of Philosophy of Science 9 (1):35-74 (2019)
  Copy   BIBTEX


In economics, thought experiments are frequently justified by the difficulty of conducting controlled experiments. They serve several functions, such as establishing causal facts, isolating tendencies, and allowing inferences from models to reality. In this paper, I argue that thought experiments served a further function in economics: facilitating the quantitative definition and measurement of the theoretical concept of utility, thereby bridging the gap between theory and statistical data. I support my argument by a case study, the “hypothetical experiments” of the Norwegian economist Ragnar Frisch (1895-1973). Frisch aimed to eliminate introspection and a subjective concept of utility from economic reasoning. At the same time, he sought behavioral foundations for economic theory that enabled quantitative reasoning. By using thought experiments to justify his set of choice axioms and facilitating the operationalization of utility, Frisch circumvented the problem of observing utility via actual experiments without eliminating the concept of utility from economic theory altogether. As such, these experiments helped Frisch to empirically support the theory’s most important results, such as the laws of demand and supply, without the input of new empirical findings. I suggest that Frisch’s experiments fulfill the main characteristics of thought experiments.

Author's Profile

Catherine Sophia Herfeld
Universität Hannover


Added to PP

896 (#15,508)

6 months
217 (#12,041)

Historical graph of downloads since first upload
This graph includes both downloads from PhilArchive and clicks on external links on PhilPapers.
How can I increase my downloads?