The mazes of logic versus the mazes of arithmetic: Keynes’s ontological commitment to the facts and events of history

Scholarship in history and philosophy of economics over the past decades largely affirmed the enduring relevance of Keynes’s concerns over the (in)completeness of theoretical model analyses, the choice of functional form, the homogeneity and measurability of empirical materials, and the use of probabilistic estimation and forecasting methods. In my contribution to this special issue, I will concentrate on Keynes’s initial concerns about the homogeneity and measurability of what Keynes quite consistently referred to as empirical ‘factors’ instead of ‘variables’. My argument is that Keynes’s concerns with econometrics were motivated by a conviction that the building blocks of sound explanations in economics did not consist of statistical data that served to quantify the (causal) variables in an equation, but of the facts and events of history, factors which are of a different order than the statistics and variables of mathematical models. Taking these factors or variables as point of departure in economic theories and explanations entail different ontological commitments, which in Keynes’s review are buried in his juxtaposition of his own ‘logical’ and Tinbergen’s ‘arithmetical’ or statistical mind. We can trace this juxtaposition to Keynes’s assessment of his illustrious predecessors Stanley Jevons and Alfred Marshall. I argue that Keynes shared Marshall’s strategy to base his analysis on the facts and events of history to reconstruct the ‘logic’ of the situation and not, as Jevons did, on the mechanisms allegedly buried in statistical data sets. This strategy sets Keynes apart from contemporaries like Tinbergen, and from the econometric revolution at large. I thus approach Keynes, from a different angle than Marchionatti (2010) and Lawson (1989), as a realist and ‘thinker of complexity’. In conclusion, I will briefly reflect on the enduring relevance of Keynes’s criticism of Tinbergen, and the pertinence of his reliance on the ‘mazes of logic.’