Statistics as a Form of Knowledge

Book cover: Columbia University Press.

In this chapter, I make the claim that the crucial form of knowledge on which they relied was statistics. A prevalent contemporary understanding of statistics equates the term with probability. In this sense, statistics have become omnipresent not only in the social sciences but also in medical research such as epidemiology, as well as in other natural sciences, and even in finance. This view is so prevalent that the main English-language monographs about the history of statistics generally focus entirely on this aspect. A second meaning of “statistics” refers to what many would call “descriptive statistics,” the “systematic collection and arrangement of numerical facts or data of any kind.” In this second sense, statistics is even more widespread. Originally, however, the word “statistics” referred to something much more specific: the description and comparison of states with quantifiable data. This is the meaning of statistics that underlies my argument.

How can such a practice ever have been novel? Nowadays, states are routinely compared on the basis of quantifiable data. The use of numbers to represent states has become ubiquitous, both in international organizations, ministries of foreign affairs, and scholarly research. But this was not always so. Some may want to argue that describing sovereigns through quantifiable facts is a practice that has always existed—for instance, in the form of the numbering of the people of Israel, Augustus’s balance sheet of the Roman Empire, Charlemagne’s inventory of his possessions, or the Doomsday book. However, none of these cases were undertaken in the spirit of modern statistics, and to think so is to “fail to understand the basis of the statistical approach’ and the ‘nature of the statistical method.” In addition to being one-off events, these instances of “counting things” were not thought to be representations of sovereigns, which could be used as templates for other sovereigns. As I will explain, such a conception and practice only emerged in the late eighteenth century.

Therefore, my first task in this chapter will be to outline when, where, and why people started systematically describing, comparing, and ranking sovereigns with numbers. In other words, when did statistics emerge as a form of knowledge as I described it in chapter 1? The chapter’s second task is to trace the transmission of statistics to the new sovereign lenders. To this end, the second section examines the dissemination of statistics in the public sphere, while the third focuses on the emergence of the real conveyor belt of statistical thinking for sovereign lenders, namely, business schools. The chapter concludes by briefly illustrating how the students trained in these new institutions brought statistics to the new joint stock banks, based on the case of the Crédit Lyonnais. In so doing, it shows that one of the outcomes of this process was the creation of systematic sovereign credit ratings.

The Development of Statistics

In a passage of his Italian Journey, Goethe describes some of the features of the town of Bolzano, where he has just arrived. He then stops abruptly and states that “in our statistically minded times” all of what he observes must already be “printed in books.” Goethe’s choice of words is interesting, as the term Statistik had only been coined in the 1740s by a German academic. And yet, as an administrator of the Duchy of Sachsen-Weimar since 1775, it is no surprise that Goethe came in contact with the emerging discipline of Statistik, developed precisely for rulers. But what was it?

Initially responsible for the rise of this discipline was a group of law and history professors engaged in the development of cameral science—the science of government—at the University of Göttingen. One of them, Gottfried Achenwall (1719–1772), coined the term Statistik in 1749. The main goal of this science, according to Achenwall and those who came after him, was to describe states’ material and nonmaterial forces, and to compare them on the basis of (mostly) quantifiable facts. The adjective “statistical” (not the noun “statistics”) had been used in earlier Italian writings such as those of Girolamo Ghilini, who relies on the adjective statistica in a monograph from 1647, mentioning an unpublished work of his on the “civile, politica, statistica, e militare scienza.” The main difference between the Italian tradition, which was concerned with collecting “information more or less systematically on foreign states,” and the German tradition is in large part that the latter was far more numerically oriented. Indeed, a comparative glance at the work of Italian writers such as Botero and Ghilini and late eighteenth-century German statisticians reveals this clear discrepancy. Beginning with Achenwall’s work, states are described and compared with quantitative data about, for example, population, debt, infantry, cavalry, and naval forces. The discipline of statistics as the practice of representing and comparing sovereigns on the basis of quantifiable facts cannot truly be said to have begun in earnest before the second half of the eighteenth century in Germany.

Though originally the matrices comparing states contained a great deal of “verbal descriptions,” this changed for the use of figures, which in turn “favored topics which lent themselves to numerical presentation.” The standard presentation became two-dimensional schemata, which had a horizontal dimension, containing the countries to be compared, and a vertical one, presenting the categories for comparison. Achenwall’s student and successor, August Ludwig von Schlözer, called for the abandonment of discursive description, as he largely preferred quantitative data. For him, statistics ought to become a “measuring discipline” and an “exact observational science” that did not describe but “contained general results as counts”—in other words, “the less adorned, the truer.” This movement took place in the second half of the eighteenth century and in the early nineteenth. Its spread, however, was not primarily caused by academic statisticians.

The increasing use of statistics by neighboring disciplines such as history and geography demonstrates how, in some sense, statistics was becoming a technique of investigation, a method of presentation of empirical facts, rather than a science in its own right. Statisticians found themselves increasingly trying to defend the remit of their discipline, which has led to the slightly confusing claim that German statistics declined in the early nineteenth century. This is a misrepresentation; though it was itself declining as a self-standing academic discipline, statistics was in fact becoming a widespread means of knowing and representing sovereigns in a variety of disciplines. Statistical descriptions of various European sovereigns started appearing in the most curious of places. It was possible, for example, to find them in the Almanac de Gotha, a German yearly almanac ranking European royalty and high nobility that enjoyed a broad audience from 1764 to the late nineteenth century. In the very late eighteenth century, statistical representations of sovereigns can be seen in a great deal of German academic scholarship, of which J. C. Gatterer and A. L. Crome’s works are good examples.

This quantitative turn to describe and know states was not to everyone’s taste. Arnold Heeren, successor and erstwhile colleague of the famous statistician August Ludwig von Schlözer, once declared that statisticians who over-relied on numbers were “table hacks.” A later commentator in the pages of the widely read Göttingische gelehrte Anzeigen (which, incidentally, was founded by Schlözer) from as late as 1807 stated in a refreshingly blunt manner: “These poor fools are spreading the crazy idea that one can understand the power of a state simply by a superficial knowledge of its population, its national income, and the number of animals nibbling in its fields… The machinations in which these criminal statistician-politicians indulge in their efforts to express everything through figures… are ridiculous and contemptible beyond all words.” Hardly a ringing endorsement of the practice, this statement illustrates the resistance with which the description of sovereigns through quantifiable facts was met, and the novelty it constituted, more than half a century after it had timidly appeared. And yet the transition from “statistics without counting” to table-statistics, oriented “largely on the methodological criteria of measurability,” was already well underway. This meant, for instance, that traditional topics such as the history of a state that were treated by the first statisticians, such as Achenwall, had to be “excluded from the sphere of interest” of Statistik.

While the heavily quantitative aspect of late eighteenth-century German statistics differentiated it from previous Italian work, it was its interstate comparison that distinguished it from the early British discipline of political arithmetic. The work of John Graunt (1620–1674), William Petty (1623–1687), and Charles Davenant (1656–1714)—central to this British tradition—was not primarily concerned with the comparison of states, but with taxes, births, deaths, and general questions of demography within specific regions or countries. Political arithmetic thus cannot be considered a form of knowledge in the sense of being an enduring way of knowing and representing sovereigns; only statistics fits this description. The Germans who picked up political arithmetic were not the international lawyers or academics developing cameral sciences at the universities of Halle, Jena, or Göttingen, but men like Johann Peter Süssmilch (1707–1767), a pastor trained in medicine, just like William Petty. This is why, when he introduced the word “statistics” to English in his Statistical Account of Scotland (1798), John Sinclair felt the need to explain that

many people were at first surprised at my using the words, Statistics and Statistical. . . . In the course of a very extensive tour, through the northern parts of Europe, which I happened to take in 1786, I found that in Germany they were engaged in a species of political inquiry to which they had given the name of Statistics. By statistical is meant in Germany an inquiry for the purpose of ascertaining the political strength of a country or questions concerning matters of state.

John Sinclair cited in Hacking, Taming of Chance, 16

This was a rather different exercise than the one proposed by political arithmeticians. One of William Petty’s treatises, Political Arithmetic, did compare England, Holland, Zealand, and France, based on a few criteria, as did Charles Davenant’s various essays concerned with France and the Dutch. However, the English development that approximates German statistics most closely was the work of a Scotsman named John Campbell. In The Present State of Europe (1752), a publication that enjoyed an international readership, Campbell presents statistical material in the form of numerical tables to assess, compare, and rank states’ power. This was part of a broader journalistic genre on the “present state of Europe,” in which a discourse identifying states as “powers” developed. Thus, the practice of describing and comparing states with numbers truly emerged across Europe in the late eighteenth century, distinctly in Germany, but in recognizably similar forms elsewhere, too.

In order to appreciate why this was the case, it is necessary to understand two developments in international relations and international political thought. First, there was the political context of the Holy Roman Empire and its constituent entities. Within this curious polity, the emperor’s power was already largely diminished by the end of the Thirty Years’ War, and even more so by the eighteenth century. While disputes between states themselves and between states and the emperor had previously been dealt with through the juridical framework of the empire, the emperor’s declining power could hardly continue to impose this old method. This became an even more acute issue in the eighteenth century with the rise of a powerful Prussia, which led to an intense rivalry with Austria. Lawyers such as Achenwall did try to maintain the old legalistic framework of the empire by omitting Prussia and Austria from their statistical work. For them, Austria and Prussia still had to be dealt with and understood within the context of the empire, not as part of an international power game.

The second important development linked to the emergence of statistics in the sense of a quantitative description and comparison of states was a new type of discourse that identified international actors as powers that were part of an “international system.” The discourse on powers and on their gradation (i.e., great, middle, or small) developed rapidly, and by the mid-eighteenth century the use of this vocabulary was “commonplace.” It was only in the latter part of the eighteenth century that the discourse on powers merged with statistics, such that gradations of power were supported by statistical analysis—for instance, in the work of Bielfeld. It is not a surprise, then, that statisticians made important contributions to debates about power and the balance of power in international relations. The idea that these European powers formed an international system, the basis of which was the freedom of each state, was most famously developed by a historian based at the University of Göttingen, Arnold Heeren, in his Handbuch der Geschichte des europäischen Staatensystems und seiner Colonien (1809). Heeren developed his argument in reaction to the French wars and Napoleon’s bid for supremacy in Europe, so as to assert that Europe had always been based on a system of free independent states. Thus, both the political context of the Holy Roman Empire and the intellectual environment in Europe were favorable to the development of statistical ideas. As an aside, it is interesting to mark the central role of the University of Göttingen in the emergence of a recognizably realist discourse on international relations organized around the notions of powers, balance of power, and international system. In the same way as we like to speak of a “Chicago school” in twentieth-century economics, or an “English school” in twentieth-century IR, there was a “Göttingen school” of international relations in the second half of the eighteenth century and the first half of the nineteenth.

Statistics was not just an obscure academic undertaking; it was eminently useful for practically oriented men, like civil servants. It was a “political discipline” meant to prepare “future practical statesmen for their duties” and impart to them the knowledge of states’ constitution and the nature of their strengths and weaknesses. The University of Göttingen, where Achenwall worked, was created in 1737 to answer the need for properly qualified statesmen. This was part of a broader movement in the development of what has been termed “cameral sciences,” or “sciences of the state,” for which professional chairs were purposefully created in various protestant German universities.

The Faculty of Law in which Achenwall worked was a key tool for dispensing education to future statesmen and administrators, attracting such promising candidates as the young Prince von Metternich, who was taught by a colleague of Achenwall’s, Johann Stephan Pütter. More minor figures trained in statistics in the eighteenth century were responsible for the great expansion of the role this discipline played in politics in the following century. For instance, Baron von Stein, the founder of the Prussian statistical bureau, studied in Göttingen and then became state minister and head of the departments of excises, customs, manufactures, and commerce. Johann Gottfried Hoffmann, head of the Prussian bureau from 1808 onward, came to play the central role in the statistical commission—the first of its kind—of the Congress of Vienna led by G. F. von Martens (another Göttingen alumnus), established after the defeat of imperial France. The commission was created in part to appease the disagreements between Metternich and Hardenberg, respectively Austrian and Prussian Chancellors, about the characteristics (e.g., population) of the territories lost by Napoleon. The commission’s work, unheard of in previous peace settlements, was key in the shift from a qualitative evaluation of state power to a quantitative one, and in the “unrestricted mathematical application of the balance principle.” Though lawyers, intellectuals, and journalists had been moving in this direction for some time, the Congress of Vienna really appears as a turning point in terms of statesmen’s practice.

Excerpted from States and the Masters of Capital by Quentin Bruneau: Copyright (c) 2023 Columbia University Press Used by arrangement with the Publisher. All rights reserved.

Quentin Bruneau is an Assistant Professor in the Department of Politics at the New School for Social Research and Eugene Lang College, and the author of States and the Masters of Capital: Sovereign Lending, Old and New (Columbia University Press, 2022).