The Statistical Method in Economics and Political Science

A Treatise on the Quantitative and Institutional Approach to Social and Industrial Problems

P. Sargant Florence

• Publisher: Routledge
• Year: 2018
• ISBN:
  9781351133494
• Edition: 1
• Topic: Commerce
• Subtopic: Commerce Général

Originally published in 1929. This balanced combination of fieldwork, statistical measurement, and realistic applications shows a synthesis of economics and political science in a conception of an organic relationship between the two sciences that involves functional analysis, institutional interpretation, and a more workmanlike approach to questions of organization such as division of labour and the control of industry.

The treatise applies the test of fact through statistical analysis to economic and political theories for the quantitative and institutional approach in solving social and industrial problems. It constructs a framework of concepts, combining both economic and political theory, to systematically produce an original statement in general terms of the principles and methods for statistical fieldwork. The separation into Parts allows selective reading for the methods of statistical measurement; the principles and fallacies of applying these measures to economic and political fields; and the resultant construction of a statistical economics and politics. Basic statistical concepts are described for application, with each method of statistical measurement illustrated with instances relevant to the economic and political theory discussed and a statistical glossary is included.

The Statistical Method in
Economics and Political Science

PART ONE

MOOD AND MATTER OF STATISTICAL ENQUIRY

CHAPTER I

The Function of Economics, Political Science, and Statistics

ECONOMICS and the adjective economic have recently become distinct, in their implication, from economy and the adjective economical. To practise economy and to have an economical outlook refers to a particular, and usually approved, “businesslike” variety of behaviour or set of opinions; to practise economics and have an economic outlook refers to the dispassionate study of all or any varieties of behaviour of a certain sort. Similarly politic like economical refers to a particular, approved, variety of behaviour, and the dispassionate study of all or any varieties is called political. What sort of matters are studied by economics and political science, and how these are related to kindred matters, will be discussed in Chapters III and IV; here it is the mood of approach, not the matter, that is at issue.

Economics and political science have as their function the study of various forms of “economic” and “political” behaviour or opinion. They may form part of an attempt deliberately to improve the behaviour of mankind, and may claim to be sciences that can be applied in certain particular ways to beneficent practical uses. But, though they must take account of other people’s opinions of right and wrong, economists and political scientists cannot alone and unaided recommend particular approved varieties of behaviour or opinion, for the reason that in the ultimate analysis they cannot judge what constitutes improvement or when a process is beneficent and worthy of approval.

Take a simple statement of economists which, whether correct or incorrect, will be intelligible to politicians:

A reduction in the population of settled countries tends to increase wealth per head.

Does this necessarily indicate what beneficent procedure to adopt? Many persons would desire larger populations for their respective countries rather than mere increase of wealth per head, and until the desideratum is agreed upon there can be no practical policy. Three distinct stages in the deliberate procedure are in fact distinguishable, each cast in an appropriate mood, the Indicative, the Optative, and the Imperative.

The Indicative is the mood of science stating a fact. A reduction in the population tends to increase wealth per head.

The Optative is the mood of ethics or morals and its ideals or desires are ultimately independent of scientific statements. They may run: Oh ! that our population were big; or else, Oh ! that wealth per head were increased.

The Imperative is the mood of the practical “economical” or “politic” statesman or publicist; but the practice deliberately commanded or advocated must depend upon the indicative statement and also upon the optative wish. It is only if the first proposition is considered true, and also if the increase of wealth per head is considered good and desirable, that the imperative may ring out,

Let the population be reduced !

These different moods can and should be separated.¹ Whatever their conduct, utterances, or opinions as private persons with the full allowance of human emotions and sentiments, there ought to be at least some authors devoting themselves to research and scientific discovery, who take their studies as seriously as do physicists, biologists, or anthropologists; and who, however keen to improve the lot of man, will be content to state their conclusions in the indicative mood and to eschew ethical valuation and practical precept. The embryonic and parlous state of economics and political science to-day, and the many controversial issues still unsolved, is primarily attributable to the admixture of moral implications with scientific observation in the author’s thought, and in the communication of his thought to reader or audience. Moral implications—the intrusion of ideals about social service and sin—obstruct and confound learning by distracting the audience, by misleading the audience, and sometimes by confusing the author himself.

The author’s moral outlook may not tally with that of the reader, and if the author mixes that outlook with his statements of fact, the latter though perfectly true, may be rejected or resisted along with the former as mere propaganda. However intelligible the author’s scientific conclusions, a “foreign” ethical code or an unfamiliar sense of sin will be a source of distraction; and communication with his audience may thus be entirely or partially baulked by his peculiar moral ideals.

Admixture with moral outlook may render statements of fact unintelligible to the audience. When some action is considered desirable and called good, or any of the fifty synonyms of good, by the author, and another action is considered a sin and called bad, or any of the hundred synonyms of bad—one simply resulting in the fulfilment, the other in the baulking of the author’s peculiar plans and desires—the reader or auditor may remain ignorant of the precise nature of the desires referred to. Probably, he will unconsciously suppose that what he (the reader or auditor) happens to consider desirable or the reverse is what the author is referring to by “good,” “bad,” etc. Here communication with his audience is vitiated by the author’s ideals and moralizings being misunderstood, and the audience is misled.

Finally, as a result of compounding ethical and moral considerations with a purely indicative attitude, the author may actually mislead and confuse himself. If definitions are based on ethical distinctions they may cut across and blur the purely factual distinctions, and if the author enunciates “principles” that might be indicative or optative or imperative in mood, he may well deduce fiction from them rather than fact.

Turning from criticism to a constructive plan, I suggest that economists and political scientists should leave the “shocked missionary stage” and imitate modern anthropological writings in their amoral attitude to social life. This attitude, applied to the solution of the problems of to-day, allows the economist and political scientist to co-operate more freely with—because independently of—practical statesmen and reformers.

This treatise, in short, assumes the function of economics and political science to be the study and discovery of fact in the indicative mood, leaving these facts open and public, to be used in conjunction with other moods for the good of mankind.

The statistical approach to Economics and Political Science must therefore be justified and valued according as it fulfils these scientific functions either in advancing research and learning or in advancing education and the power to learn.

The phrase that rankles most among statisticians ambitious to advance learning and research is that statistics prove nothing, but “merely describe.” What is implied by this term of contempt ? On the most severe interpretation “merely descriptive” means descriptive merely of an isolated fact. Some individual perhaps lived to be 150 years old, or Blue Books may tell us that the population of London in 1901 was 4,536,063 souls. But statistics, though made up of this sort of thing, are not this, any more than a house is merely a set of bricks. Populations of single towns and individual’s ages are the data or “item cases”; but vital statistics do not begin till rates and averages are calculated, nor population statistics till curves of growth and other comparisons are instituted. The essence of the statistical approach is summarization of the numerical data, the measurement of their general value or range of value, and of their degree of variation; and if by description is meant the mere recording of isolated and unrelated facts, the inconsequent retailing of anecdote, then statistics is not mere description.

“Mere description” may, however, be given a wider reference. It may, while excluding generalization and inference, yet include this numerical summarization that reduces masses of item cases and stray information to manageable proportions.

Even in this sense of description, works of “mere” description are often dismissed as unworthy of the attention and beneath the consideration of thinking persons. Yet under certain circumstances, I venture to submit, a wider knowledge of the facts such as statistical summaries would give is the one thing needful to change merely thinking persons into scientific investigators. These circumstances are all too often present in the case of English economists and political scientists.

There are certain summaries of facts such as the form of the actual distribution of wealth between classes of individuals, or the standard types of consumption of various social classes, which economists should not merely take under notice by consulting some reference book or otherwise refreshing their memories, but which they must all the time keep at the back of their heads or at least “paste in their hats.” Otherwise things, often inaccurate things like equality of opportunity, will be too much taken for granted. In England taking such things for granted is peculiarly dangerous owing to the odd method of educating the class from which most economists and political scientists are drawn. Boys are deliberately segregated from girls and from their own families for eight months in the year, and owing to the expense of this education they are automatically kept apart from the sons of poorer parents. At the most impressionable ages they obtain no first-hand experience of the behaviour of half their own class or of any member of what they come to consider a lower class but which, be it remembered, constitutes at least four-fifths of England’s population. Unless they have family and working-class problems and conditions described, middle-class young men may well remain in ignorance; and should they proceed to argue deductively, introspectively, and from limited premises “out of their heads,” they may not only delude themselves but lead others astray.

English classical economists preached thrift unceasingly to the working-classes in blissful ignorance of the probability (revealed in vital statistics) that savings of earnings would result in curtailment of bare necessaries in the way of food, housing, or education. English classical economists were also responsible for the idea that birth rates depended on married men’s ratiocination as to their power of supporting additions to the family, and that in consequence birth rates would vary positively with increases in wages and other forms of income. The bare statistical description of the course of birth rates in the different classes since 1870 has been sufficient to dispose of this.

A further characteristic of the primarily classical education of the well-to-do Englishman is its attempt to make him a gentleman who shall regard a life of business activity as materialistic, sordid, and somewhat vulgar. “The average person is not interested in Industry” was the first sentence of an essay turned in to me by one fresh from a “public school.” It is not surprising that Mr. Gordon Self ridge has a hard task in getting the “romance” idea of business across the English footlights.

While the perfect English gentleman should not go into business, the upright American citizen will not touch politics. Hence in one country or the other economic and political behaviour tends to be studied as an exercise in dialectic, to the neglect of the actual modus operandi of economic and political institutions. The behaviour of joint-stock companies and corporations, of boards of directors and factory managers, or of committees of municipal councils is left severely alone, though a statistically summarized description of their practice and experience would correct many assumptions taken for granted, and shift economics and political science from a merely speculative and rhetorical basis to a basis of fact.

Here I rest the case for mere description in the sense of summarized information about specific events. The claims of statistical method for a further extension of its frontier in the world of science are more controversial. “The theory of statistics,” writes J. M. Keynes,¹ “can be divided into two parts which are for many purposes better kept distinct. The first function of the theory is purely descriptive. It devises numerical and diagrammatic methods by which certain salient characteristics of large groups of phenomena can be briefly described; and it provides formulae by the aid of which we can measure or summarize the variations in some particular character which we have observed over a long series of events or instances. The second function of the theory is inductive. It seeks to extend its description of certain characteristics of observed events to the corresponding c...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Preface
• Contents
• List of Tables
• Part One Mood and Matter of Statistical Enquiry
• PART TWO THE STATISTICAL MEASUREMENT OF ECONOMIC AND POLITICAL DATA
• PART THREE STATISTICAL FIELDWORK IN ECONOMICS AND POLITICAL SCIENCE
• PART FOUR SKETCH FOR A STATISTICAL ECONOMICS AND POLITICS
• Glossary of Elementary Statistical Terms
• Alphabetical Index