Economic experiments: between science and ideology


Institutional Communication Service

6 May 2020

The experimental approach to scientific research is a fundamental element in the vast world of natural and technical sciences. Social sciences, including economics, also use the experimental approach. In a contribution recently published in the economic and financial newspaper Finanz und Wirtschaft, François Degeorge, professor of finance at USI and director of the Swiss Finance Institute (SFI), compares scientific and ideological experiments.

In economics the experimental approach is well illustrated by the work of the three recent Nobel Prize winners, Abhijit Banerjee and Esther Duflo of the Massachusetts Institute of Technology (MIT), and Michael Kremer of Harvard University. “These scholars have carried out experimental work at the forefront of the fight against poverty, studying the impact of various public policies on the economic prosperity of emerging countries", explains Prof. Degeorge. The work of these researchers shows that the experimental approach requires a high degree of coherence and discipline. Whether an experiment is judged a success or a failure depends without exception on the outcome of the experiment itself. The criteria for a possible success are precisely defined at the beginning of the experiment - also in order to keep as little margin for interpretation as possible after an experiment fails.

Certain economic policy decisions aimed at resolving crisis situations can also be described as economic experiments. One example is the reaction of central banks to the 2008 financial crisis. "These economic policy experiments, however, are fundamentally different from scientific experiments because they usually aim at the rapid resolution of urgent problems, and not at the validation of theories. Moreover, even though economic policy experiments are sometimes necessary, the knowledge gain they bring is usually modest,' Degeorge points out.

A third type of economic experiments is purely ideological. This type of experiments offers neither solutions to problems nor progress in knowledge. For their promoters the criterion of success is political. They declare victory as long as the experiment is merely carried out, regardless of its consequences. "With the recently proposed initiative for a micro-tax on all cashless monetary transactions in Switzerland, we have a typical example of an ideological experiment. The proponents of this initiative make great promises and offer simplistic solutions to complex challenges. This initiative is intended to replace three federal taxes at once - VAT, direct federal tax and stamp duty - with a new model. In doing so, the proponents of the initiative exclude from the outset the possibility that their experiment might fail. Ideologically motivated experiments clearly contrast with the well-founded, data-based approach that characterises established economic science," says Prof. Degeorge.


The complete article in German is available online here: