STATISTICS is that department of political science which is concerned in collecting and arranging facts illustrative of the condition and resources of a state. To reason upon such facts and to draw conclusions from them is not within the province of statistics; but is the business of the statesman and of the political economist.
That it is necessary for a government, in order to govern well, to acquire infor mation upon matters affecting the condi tion and interests of the people is obvious. Indeed, the civilization of a country may almost be measured by the completeness of its statistics ; for where valuable statis tical records of antient date are found concerning a country not yet advanced in civilization, which would appear to con tradict this position, we owe them to sovereigns or governments of uncommon vigour and sagacity. However rude the government of a country may be, it can not attempt to make laws without having acquired the means of forming a judg ment, however imperfect, as to the mat ters brought under its consideration. In this sense statistics may be said to be coeval with legislation ; but as legislation has rarely been conducted upon any fixed principles, or partaken of the character of science, in the earlier ages of the world, we must attribute to statistics, as a department of political science, a much later origin. It is chiefly to the rise of political economy that we are indebted for the cultivation of statistics. The principles of that science, which are di rectly concerned about the prosperity and happiness of mankind, were not reduced to any system until the middle of the last century : since that time, political eco nomy has been cultivated as an inductive science. The correctness of preconceived theories has been tested by the observa tion and analysis of facts ; and new prin ciples have been discovered and esta blished by the same means. A limited knowledge of facts had previously been an obstacle to the progress of political economy ; and, on the other hand, the neglect of that science caused indifference to statistical inquiries. Statistics, which had been neglected until political eco nomy rose into favour, have since been cultivated with continually increasing care and method, as that science has been further developed, and the knowledge of its fundamental principles more widely diffused.
This connection between political theo ries and statistics, while it has led to the collection of many data which would not otherwise have been obtained, has often introduced a partial and deceptive state ment of facts, in order to support precon ceived opinions. This is sometimes un justly objected to statistics, as if it were a defect peculiar to them. That facilities for deception are afforded by statistics cannot be denied ; but fallacies of this kind, like all others, are open to scrutiny and exposure. Reliance need not be placed upon statements of facts nor on numbers, unless supported by evidence ; and inferences from them should only be admitted according to the rules by which all sound reasoning is governed. Falla cies are difficult to detect in proportion to the ingenuity of the sophist and the igno rance or inexpertness of his opponents ; but in political matters, opposite theories and opinions are maintained with equal ability, and facts and arguments are in vestigated with so much jealousy, that, in the end, truth can hardly fail to be esta blished. Neither does any suspicion of partiality attach to such facts as are col lected by a government without reference to particular theories. Until some one has shown the valve of noting a certain class of facts with a view to his own inquiries, no pains are taken to obtain information of that nature from the best sources; but as soon as the importance of seeking any data is acknowledged, the collection of them becomes the business ce impartial persons. The statist must be acquainted with the purposes to which the facts collected and arranged by him we likely to be applied, in order that the proper distinctions and details may be noted in such a manner as to give the fullest means of analysis and inference ; but his services are greatest when he does not labour in support of a theory.