« 上一頁繼續 »
• Population,' says M. Quetelet, is the statistical element, par excellence: it necessarily rules all others, since it relates, above all, to the people and the appreciation of their welfare and their wants. It would be vain to attempt to form statistics of value without taking as a basis the results of a census executed with all the care and precision which so delicate an operation requires. The other data have no real value, except in so far as they relate to the number of the population. A census carefully made sums, in a measure, the most important problems which can be proposed to a statist. The classification according to age allows of the establishment of tables of population, of forming correct ideas on mortality, on the forces at the disposal of the state in case of necessity, and of fixing the ratio between the useful fraction which contributes to the general wellbeing, and the fraction which yet requires assistance and support to become in its turn useful. The classification by professions, indicates the means by which the population provides for its subsistence and tends to augment its prosperity. ... Those by civil condition, by origin, by education, furnish the administration with no less precious information to assure internal good order, and to facilitate the execution of the laws. (Transl. p. 183.)
A well-organised system of civil registration, (' état civil,') is therefore one of the first wants of an enlightened people. No man in such a people is above or beneath the obligation of authenticating his existence, his claims on the protection of his country, and his fulfilment of the duties of a citizen, - or of contributing bis individual quota of information, in what personally concerns himself or his family, in reply to any system of queries which the Government in its wisdom may see fit to institute respecting them. Such information inay be regarded as a poll-tax, which, in this form, a Government is fairly entitled to impose, and which indeed is at once the justest and least onerous of taxes; or rather, it may be looked on as a mode of self-representation, by which each individual takes a part in directing the views of the legislature in objects of universal concern. Nothing, therefore, can be more unreasonable than to exclaim against it, or to endeavour to thwart the views of Government in establishing such a system, -- nor anything more just than to guarantee its fidelity by penalties imposed on false returns or wilful omissions.
The analysis of the population returns of a great nation, or rather the drawing from that analysis, duly executed according to rational classifications, just and philosophical conclusions, is a task calling for the exercise of much acuteness and discrimination in appreciating the influence which the relative proportions between the classes, as to age, condition, calling, must necessarily have on national character and habits, and in weigh
ing— with reference to future prospects — the probable influence on that character and those habits which is involved in even a very moderate observed change from time to time, in those proportions.
• The numerical tables of a population, when made with care and with all the developement which science requires ... form, in the annals of a people, the most eloquent page that a statesman can read, if he understand them well. In fact it only belongs to the practical observer completely to understand the language of figures, and not to go beyond what they can teach him. Censuses, well made, and which succeed one another on a uniform plan and at intervals suffi. ciently near, should present most precise notions of the physical and moral condition of a people,- of the degree of its power, – of its prosperity,—and of the tendencies which may compromise its future: they would teach much better than voluminous inquiries, which are often fettered by prejudices and private interests, what we ought to think of the retrograde state or the immoderate developement of certain branches of industry.'
Among the first results of such an analysis, are those general ones which our Continental neighbours technically understand by the movement of the population—its increase, that is to say, by the excess of births over deaths and emigrations, and the internal change in the proportions of those living at different ages corresponding to changes, if any, in the law of mortality as indicated by the ages of death. On this point M. Quetelet, in an earlier part of this work, makes the following pertinent remark.
• The movement of a stationary population is often compared with that of a population increasing by an excess of births over deaths. However, this is a comparison of heterogeneous elements : all other things being equal, the latter population should have a greater mortality; for there are more children in it.'
So far as this remark goes it is just, but it does not include the whole case, or exhibit fully the influence of the consideration in question. To judge of the extent of this influence it is only necessary to consider that, in a given population now existing, the individuals living at any assigned age are not the survivors of that age among a number equal to that born in the current year, but among a number born antecedently, when the population was less than at present, in a proportion easily calculated, the age being given, and the annual rate of increase known. Thus, supposing the population of a country to double in fifty years; a man fifty years old is the survivor of only half the number of cotemporary births, and of one hundred of only onefourth those which would appear, on a comparison of the number actually born in a given year with those actually living at the
age specified, in that year. Not only, therefore, are there more children in comparison with adults in an advancing population, but at the same time fewer old men. Now the ratios of the helpless, the active, and the meditative elements of a population to the entire mass and to each other, -of giddy youth and adult enterprise to mature experience, timid caution, and declining powers, must necessarily give rise to corresponding features of national character. A disproportion in this respect, influencing all the great lines of developement of national activity and impressing the whole career of a people, cannot but make itself felt in every feature of their existence. It is only necessary to contrast the energy displayed by a nation whose population doubles in twenty-five years, as in the United States, with the sobriety of movement, not to say torpor, of another, where, as in Holland, it is nearly stationary, to perceive the connexion in question to be that of effect with cause.
• An exposition of the political condition belongs essentially to the statistics of a country. We do not, however, know how to express it in figures. The same may be said of information relative to the moral and intellectual condition. The simple recital of what has passed in a locality at a particular time sometimes better teaches the moral condition of a people than all the numerical tables possible.'
Statistics, however, deals essentially with numbers. It may be difficult, or impossible, to express numerically the degree of political freedom, the extent to which the institutions of a country fulfil the ends of their establishment and maintenance, or the degree in which its fiscal regulations press upon its inhabitants, — yet these are nevertheless results capable of being estimated, and which it is of the last importance to estimate; and the estimation must ultimately rely, to a considerable extent, on the numerical exhibition of particulars. Thus, to say nothing of the statistics of elections in which numbers are easily and precisely attainable, or of those of crime; accurate returns may and ought to be obtained and published of a great variety of particulars relative to the administration of justice in our civil courts, by which our judgment as to their well or ill working may be influenced. As examples, we may specify the statistics of juries, common and special, – those of legal decisions in civil cases, more especially as regards the cases of new trials moved for and obtained, and their grounds; - of decisions appealed from to higher tribunals, and of the proportion of cases in which such new trials or such appeals have affirmed or reversed the former decision, - points of great interest as concerns the confidence with which the decision of a civil court may be relied on by its suitors, but of which, if any official
the originalt to law. Seteculiar circumstanuired to p
returns exist in this country, we have been unable, after some considerable amount of inquiry, to procure them. This is to be regretted, because the application of the theory of probabilities to judicial decisions with this very view (that of determining, from the amount of self-contradiction existing among them, their value as tests of truth,) has been expanded by Laplace and Poisson into a very elaborate theory, which the latter especially has applied to the statistical returns of the French tribunals, civil as well as criminal. It may be worth while here to mention the conclusions deduced by the last-mentioned geometer from the consideration of 17157 cases adjudicated on in French courts of civil appeal, during the years 1831, 1832, 1833. Of these the judgment of the inferior tribunal was confirmed in 11747 cases, or in 685 cases out of a 1000,- a percentage certainly not calculated to inspire a high degree of primâ facie confidence in the efficacy of a resort to a court of justice for the redress of a civil injury, though it must be admitted that appeals would chiefly take place in cases where the original decision was obviously contrary to common sense at least, if not to law. Setting out with these data, and taking into consideration the peculiar circumstances of the French institutions, in which three judges are required to pronounce, by a majority of voices, a jugement de première instance,' and seven in a court of appeal, M. Poisson concludes the probability that a confirmatory decision will be a just one, to be 0-948, or about 19 to 1 in its favour, and 0.6+1, or about 16 to 9 that a reversal of the former decision will be so. With respect to the probability that a second appeal will confirm the decision of a previous one, be that in favour or not of the original decision, he assigns it at 0-7466, or about 3 to 1 in favour of its doing so.*
Taxation, too, is an element of political condition easily enough represented in figures, but in which it is hardly possible to get two persons to agree in their interpretation. M. Quetelet sums up his few and cursory remarks on this subject with a dictum which, notwithstanding Mr. Norman, most Englishmen will feel to be intended for their peculiar consolation :- It has • been justly remarked,' he says, that those are the most civilised • countries who (which] pay proportionally the most to the government.
The chief difficulty to be encountered in aiming at correct results in the collection of agricultural, industrial, and commercial statistics is, that it
*Requires the intervention of persons who are almost always
Paris : 1837.
* Recherches sur la Probabilité des Jugemens. Della Autorità giudiciaria, di F. Sclopis, p. 84.
interested, or think they have an interest, in disguising the truth. When the government collects them, it is generally opposed by the manufacturer, who supposes it done with fiscal views. The desire to obtain freedom for his industry, and to obtain what are called protecting laws ... almost always tends to exaggeration in one direction or another. Governments also publish documents on importations and exportations. These tables, which are useful to consult, nevertheless often contain very vague returns : they are generally confined either to the fixing of prices from faulty valuations or of quantities without considering either price or quality. In the official valuations, moreover, we only know a part of the truth : it is especially here that information not susceptible of reduction to numbers becomes necessary, in order to determine the probable quantity which escapes the legally stated values.'
Owing to these causes of jealousy and incomplete presentation, many important statistical elements, relating to matters of pecuniary concern, can hardly be collected by official intervention. It is here that a Statistical Society may render most valuable service by setting on foot systematically, yet amicably and unobtrusively, local and private inquiries, with the guarantee of personal veracity for their answers, and the purely scientific and truth-loving spirit of such a body of enlightened inquirers for their fair presentment.
• The statistics of the moral and intellectual condition of a people,' he goes on to observe, present still greater difficulties; for the appreciation can only be founded on facts much more contestable than those given by industry and commerce. When we say that a province produces so many quarters of corn or so many gallons of oil, we know that the figures may be more or less in error; but we understand the nature of the unit. It is not the same when we say that a province produces annually so much crime..... Infinite precaution and sagacity are necessary to read with success the statistics of tribunals, for the documents they contain are very complex in their nature, and almost always incomplete.' ....
• What a mass of errors have we not accumulated in treating of pauperism! To probe this leprosy of society we have had recourse to lists of the poor, and very often without inquiring if these lists were complete and comparable in different countries or even within the limits of the same country. Real poverty is nearly always very different from the poverty officially returned. .... In Belgium a man will enter his name on the list of paupers to escape serving in the civic guard, or to obtain other advantages, without receiving a farthing of public benevolence '[!!]
With such difficulties in the way of exhibiting fairly, and interpreting truly, statistical facts, arises a necessity for laying down precautionary rules for the guidance of those to whom is confided the important task of their collection and registry — for checking their correctness when collected — and for their legiti