Statistics, that branch of Political Science which has for its object the collecting and arranging of facts bearing on the condition, social, moral, and material, of a people. The collecting of such facts, and the taking of censuses for military purposes, have been in use since the earliest times: King David numbered his people, and Egyptians and Romans had censuses. But the treatment of the statistics of all nations as a branch of study dates from the time of Conring (1606-81), in Germany, to whose distinguished successor Achenwall of Göttingen (1719-72) the name of the study (Ger. Die Statistik) seems to be due. As distinguished from the early and simple 'descriptive statistics,' in which the figures were but illustrations to the text, a more scientific arithmetical or mathematical method may be credited to Sussmilch (1707-67), whose work had been simplified by the working out of probabilities and averages in connection with mortality tables and otherwise, by Petty and Halley in England, and others. But most of all to Quételet (q.v.), the great Belgian statistician, is the science indebted for its present standing. The principle lying at the foundation of the science as ultimately developed is that the laws which govern nature, and more especially those which govern the moral and physical condition of mankind, are constant, and are to be discovered by the investigation and comparison of phenomena extending over a very large number of instances. Accidental diversities tend to neutralise each other, their influence diminishing as the area of investigation increases; and if that area be sufficiently extended, they so nearly disappear that we are entitled to disregard them altogether. While the length of a single life cannot be counted on, an average of 1000 or 10,000 lives gives us a constant quantity, sufficiently near the truth to answer the purposes of insurance companies. Even the acts which are the most purely voluntary as regards individual men have been found to be subject to laws which, in respect of the masses which make up society, are invariable in like circumstances, and discoverable.
The science of statistics has a twofold relation to political and social economy. The facts collected by the statist are the bases on which political economy rests; their application to social and economical problems is an appeal from imagination to fact. But the statist must be guided by the political economist in what direction to extend his investigations: without political economy we should have had no statistics. There have been keen and useless controversies as to whether statistics is an adjunct to other sciences, and a mere method, or an independent science; and as to whether it should limit its scope to national and social phenomena (in which acceptance the word 'demography' has been proposed as a descriptive name), or should be extended into meteorology and other natural sciences. It is impossible to give any convenient and comprehensive classification of the multifarious topics that fall within the sphere of the statistician—population, trade inland and foreign, wealth, currency, prices, banks, social conditions, the people, &c.
There was a kind of statistical bureau in France in Sully's days; such an institution was permanently set agoing in 1800. But the perfecting of statistical methods owes more to the foundation of the Belgian statistical bureau under Quételet in 1831. Since then most civilised lands have devoted much time, labour, and money to collecting and tabulating their statistics—France, Italy, and Germany, and the United States being perhaps remarkable for the fullness and systematic organisation of their statistical returns. What is done in England is not managed by any one central bureau or board, and is less systematic, though very varied and valuable (see BLUE-BOOKS).
There was a statistical section added to the British Association in 1833, and the Statistical Society of London was established in 1834. Under Quételet's influence a great statistical congress was brought together at Brussels in 1853, and like congresses have been held since, usually at intervals of three years, in one of the chief European towns. The Journal of the Statistical Society and the Journal de la Société de Statistique of Paris appear regularly; and in 1885 an 'International Institute of Statistics' was founded, which publishes a Bulletin de l'Institut Internationale de Statistique.
See the articles in this work on CENSUS, MORTALITY (BILLS OF), INSURANCE, GRAPHIC METHODS, AVERAGE, PROBABILITIES, VITAL STATISTICS; the relevant sections in the articles on Great Britain and the several countries, as also such articles as COTTON, RAILWAYS; annuals like the Almanac de Gotha, Statesman's Year-Book, and such almanacs as Whitaker's; Mulhall's Dictionary of Statistics (1883; new ed. 1891-92); Kolb's Condition of Nations
(Eng. trans. 1880); Hübner's Statistische Tafel (annual); Webster's Trade of the World (1880); Block's Traité Théorique et Pratique de Statistique (1878); and other works and articles about statistics by modern statisticians like Bodio, Haushofer, Kries, Gabaglio, Farr, Giffen, and Sir Rawson Rawson.