A History of Econometrics in France

Philippe Le Gall

• Éditeur: Routledge
• Année: 2007
• ISBN:
  9781134352555
• Édition: 1
• Sujet: Business
• Sous-sujet: Business generale

This text challenges the traditional view of the history of econometrics and provides a more complete story. In doing so, the book sheds light on the hitherto under-researched contribution of French thinkers to econometrics. Fascinating and authoritative, it is a comprehensive overview of what went on to be one of the defining subsets within t

Part I.
Natural econometrics and the “system of the world”

De toutes les sciences naturelles, l’Astronomie est celle qui présente la plus longue suite de découvertes. Il y a extrêmement loin, de la première vue du ciel, à la vue générale par laquelle on embrasse aujourd’hui, les états passés et futurs du systême du monde. Pour y parvenir, il a fallu observer les astres, pendant un grand nombre de siècles; reconnoître dans leurs apparences, les mouvemens réels de la terre; s’élever aux loix des mouvemens planétaires, et de ces loix, au principe de la pesanteur universelle; redescendre enfin, de ce principe, à l’explication complète de tous les phénomènes célestes, jusque dans leurs moindres détails. Voilà ce que l’esprit humain a fait dans l’astronomie. L’exposition de ces découvertes, et de la manière la plus simple dont elles ont pu naître les unes des autres, aura le double avantage d’offrir un grand ensemble de vérités importantes, et la vraie méthode qu’il faut suivre dans la recherche des loix de la nature.

– Pierre-Simon Laplace, Exposition du Systême du Monde.

The next three chapters are explorations into an unfamiliar region of the history of economic ideas: the rise and the development of natural econometrics in France, from the 1830s to the 1860s. Despite differences, for instance relative to the way they use mathematics,¹ Antoine Augustin Cournot, Jean-Edmond Briaune, and Jules Regnault share much at the methodological and the philosophical levels. All three drew inspiration from scientific programs of the period (mainly those defined by Laplace and Quételet), and analyzed economic issues with the help of statistics (data and instruments) and mathematics. Such a use of statistics and mathematics is closely dependent on their image of the economy and the society: both would be made up with causal, deterministic, divine and mathematical laws and structures (most often of a mechanical nature), and our three authors conceived statistics as a means to discover such laws and structures. In addition, these laws were understood as being parts of a general natural order, which would rule the social world as exactly as the natural world – the whole world was considered mathematically and mechanically organized and structured, and originating in a divine design. In that perspective, Cournot, Briaune, and Regnault aimed at discovering the secrets of the “system of the world” and at making Nature confess.

In the hands of these scholars, observation and statistical instruments play a central role. For Cournot, Briaune, and Regnault, science originates in observation. To my knowledge, none of them was directly involved in the collection of measures. But “from about 1830,” as Theodore Porter shows, “there occurred an enormous increase in the acquisition and use of quantitative information about nature, technology, and society alike” (Porter 2001:14). They could thus use data collected by administrations and institutions, for instance the Ministry of Agriculture or Paris Stock Exchange. Yet, for all three observed data were systematically associated with imperfection. Imperfection does not refer here to a bad quality of observed data, but rather to the functioning of the social world: although it is not chaos, this world would incorporate perturbation – for instance historical accidents or dissenting behaviors originating in ignorance – or, more radically, it is seen as having not reached yet its final and harmonious form. This explains that for these authors, statistical instruments had to be used in order to unveil the signs of constancy, order, and perfection in data. Here the statistical instrument par excellence is the average – these authors share much with Quételet in that perspective² – from which accidents and imperfection could be eliminated. From averages, these scholars thought that timeless, invariant laws could be inferred and discovered: the “law of sales,” which relates price and quantity (Cournot); the “laws of proportionality,” which connect price to supplying (Briaune); and the “law of attraction” governing stock price behavior (Regnault). It is at this stage that statistics and mathematics become inseparable. From observation and from the use of statistical instruments, all three thought that it could be possible to identify mathematical laws ruling the society. In other words, for natural econometricians statistics plays an active role: it was considered a means to discover mathematical relationships and laws. Indeed, Cournot, Briaune, and Regnault believed that the social world – just like the natural world – would be a text written in mathematics, a text that it could be possible to decipher with the help of appropriate statistical instruments. In sum, we shall see in the next three chapters that for these three authors, the use of averages was seen as a means to unveil mathematical laws that would govern the society.

We can perceive here the way the methodology of natural econometricians and an idiosyncratic worldview are closely intertwined. These authors rooted their explorations of the economy in the belief that the social world would be, just like the natural world, ruled by a natural order. As we shall see in a more detailed way in the next chapters, the use of statistics and mathematics is here based on a particular image of the society. First, the society would be governed by a finite set of mathematical and mechanical³ laws. This idea legitimates two kinds of unity: an ontological unity – the whole “system of the world” would be made up with a set of laws that apply to society and nature – and a methodological unity – this worldview explains these authors’ search for economic translations of laws at work in the natural sciences. In other words, reductionism is here legitimated by the unity of phenomena supposed to prevail in the real world. Second, this ontological unity would find its roots in God’s design. In fine, natural econometricians saw statistics and mathematics as means for knowing laws they considered divine laws, and such laws were positively interpreted: they would work for harmony, at the individual and at the collective levels. For them, the social scientist would be an observer and a discoverer of God’s secret workshop, of a world they consider finite and harmonious. This explains a pivotal feature of natural econometrics: the task of the natural econometrician is to discover in observed data such a hidden order that, once identified, could lead societies to a stable and harmonious state, in full conformity with natural order and God’s design. Not surprisingly, for Cournot, Briaune, and Regnault, statistics and mathematics were means for identifying an “end of history,” the moment when no deviations from the natural laws occur and when historical time has no room anymore, when the economy becomes a timeless, ahistorical system.

The next three chapters are illustrations of this idiosyncratic interweaving of econometric ideas and a worldview dominated by the belief in a divine natural order. However, it is important to keep in mind that although econometric ideas emerged and took shape in the hands of Cournot, Briaune, and Regnault, this emergence never led to the constitution of a disciplinary field at the institutional level: natural econometrics remains a story of rather isolated figures. Basically, these authors can be considered innovators, in the sense that they largely break with the common practice and methodology of political economy of their time. Yet, they also perfectly fit into scientific paradigms of the ...