CENSUS is now almost solely used to denote that enumeration of the people made at intervals in most European countries, and in the United Kingdom and the British Colonies decennially. The term had its origin in Rome, where a group of the many functions performed by the high officer called censor received the name of census. An enumeration of the people was only one of them, but they were chiefly of a statistical character. They were especially directed to fiscal objects; and it does not appear that the enumeration of the people was then deemed of value as a source of statistical knowledge which might influence morals and legislation. It was connected with the Servian constitution, which apportioned the rights and duties of citizens to the amount of property, dividing them into six classes, which were subdivided into centuries by a mixed ratio of wealth and numbers. Had the enumeration been deemed of value for any such other purposes, besides the adjustment of rights and obligations, as those for which statistical knowledge is now deemed so valuable, the notices preserved of the vast collection of statistical facts thus made would have been less scanty and meagre, and we should not have found it so impracticable to come to any conclusion about the population and extent of the city of Rome itself. The Roman census must have been minute and full. It indicated not only the number and respective classes of all free persons, but their domestic position as husbands and wives, fathers and mothers, and sons and daughters. The slaves and freedmen were indicated in connection with the possessions of the head of the house, and landed property was analyzed into several classes according to its character and produce. The important practical effect of the census caused it to be conducted at intervals generally so frequent as every fifth year. It was followed by a sacrifice of purification or lustration, whence the term of five years came to be designated a lustru m. There were highly penal consequences to the citizen who neglected his registration for the census, to whom as an unregistered person the name of incensus was given. From the mixed functions to which it was applied we have the word used among the Romans to signify the patrimony or property qualification of a particular grade—as census senatorius and census equester; and we have it employed in later times to indicate taxation. Hence census dominicains, implying a feudal tax to the superior, and census duplicatus, a double tax or feudal casualty; and the word cense, used by old English writers, was abbreviated in modern use into cess.
While the word census was thus applied to the taxation of the Middle Ages, it will readily be understood that in its modern sense it received no practical application, since neither taxation nor the adjustment of social rank required a numbering of the people; and the statistic or economic ends of such a process were as little known as they had been to the Romans. Under the despotic Governments of the Continent, however, the tendency to central organization for purposes of administration and police prepared the way for statistical inquiries into the numbers of the inhabitants of particular areas whenever there should occur an occasion for enumerating them. It was in Britain, with its abstinent Government and unrestrained people, that the want of population statistics became most flagrantly conspicuous. It is difficult at present to realize the idea that, long after Adam Smith’s time, the number of the inhabitants of the British empire could only be guessed at as the populousness of China is at the present day; and, as in all matters of statistics, which have their own simple solution through specific inquiry, the guesses about the population of the empire were not only vague but extravagantly contradictory. During the 18th century, the most trustworthy geographers were generally those who did not venture on an estimate of the population even in those European states which had the best means of enumeration at their command.
The first effort to take a census of the population of Great Britain was made in 1801; it did not then extend to Ireland. The success which attended this and the two succeeding efforts was mainly owing to the zeal and ability of Mr Rickman, the assistant-clerk of the House of Commons. Where there is an organization like that of many in the European states for preserving a constant official record of all the fluctuations of the population, not only in their absolute numbers throughout a whole territory, but in the relative numbers in its respective parts as they may be affected by fluctuations, systematic arrangements are thus prepared not only for obtaining a general census at any one moment, but for checking its accuracy and classifying its elements. But to deal at once with the raw material in the self-governed British empire required great ingenuity and sagacity. A census, to be accurate, must be taken on a uniform system, and must be taken simultaneously. Any enumeration going over a tract of time, were it but two days, must be more or less inaccurate, and destitute of the means of correcting its own inaccuracies. Besides the mere abstract numbers of the people, there is much collateral information to be recorded. This, besides its own intrinsic value, is necessary as a check on the numbers; since a distribution into elements according to sex, age, social condition, occupation, and the like, affords a self-acting control on the accuracy of mere figures. In a census, indeed, it is a simple rule, that the information returned should be extended as far beyond the main facts as with safety to these it can be carried. The tendency towards complexity in the nature of the returns must always be checked by the liability of the people at large to make blunders and create confusion where they are required to attest facts not of the most obvious nature, and by the difficulty of getting a number of subordinate officers to understand and carry out a complex classification. Hence there was great difficulty in obtaining a classification according to occupation, from its complex intermixture with the classification according to families. Thus, in the first census, there was an attempt to classify the people under three divisions—(1) persons chiefly employed in agriculture, (2) persons chiefly employed in trade, manufactures, or handicraft, and (3) all other persons not comprised in the two preceding classes. But Mr Rickman found the returns unsatisfactory, from the difficulty of deciding “whether the females of the family, children, and servants were to be classed as of no occupation, or of the occupation of the adult males of the family.” In the two subsequent enumerations, the rule adopted was to record the occupation of the head of the family; but here comes a new element of confusion, in the difficulty of defining the head of a family. Experience, and an anxious desire to combine simplicity and comprehensiveness in the returns, were the only means by which such difficulties could be mitigated.
The enumerations of 1841 and 1851 in England were much facilitated by the uniform system of registration of births, marriages, and deaths which came into operation on 1st July 1837, and which not only afforded the means of checking the accuracy of the returns, but provided a prompt and skilled machinery accustomed to statistical work. Far more dependence could now be placed on the discretion and skill of the officers to whom the local duties were committed; and the returns were made more minute and complete. In Scotland it was necessary to adopt the method of employing the parish schoolmasters to perform the local duty of enumerating the population in the country districts. In Ireland the first attempt at a générai census was made in 1811, but it was decidedly unsuccessful. It was repeated in 1821, but went no further than a bare enumeration of doubtful accuracy. The census taken there in 1831 was subject to correction in 1834, to make it the basis of the new system of national education. In the two subsequent enumerations the aid of the admirable constabulary force, and the use of an ordnance survey, nearly complete in 1841, went far to supply the want of permanent local statistical machinery.
The census of 1851 was taken on the 31st day of March, the previously distributed schedules beiug then collected. They embraced a return of the local and other conditions of the population during the preceding night.
“At the present census,” say the commissioners, “it was resolved to exhibit, not merely the statistics, as before, of parishes, and more completely of parliamentary and municipal boroughs, but also of such other large towns in England and Scotland as appeared sufficiently important for separate mention, and of all the ecclesiastical districts and new ecclesiastical parishes which, under the provision of various Acts of Parliament, have, during the last forty years, been created in England and Wales. In addition also to the inquiry concerning the occupation, age, and birthplace of the population, it was determined to ascertain the various relationships (such as husband, wife, son, daughter), the civil condition (as married, unmarried, widower, or widow), and the numberof persons blind, deaf, and dumb.”
Another novel feature in the census of 1851 was an attempt to supply the statistics of the ecclesiastical and educational condition of the country. It stated the amount of church accommodation at the command of each religious denomination; while a return was procured of those in attendance in the several churches on Sunday, 30th March. The attendance throughout thirty-five religious communities in England on that forenoon was returned as 4,428,338, of which the proportion assigned to the Church of England was 2,371,732. The returns for Scotland, admittedly very imperfect, give a total of 943 y 951, of which 351,454 belonged to the Established Church. The English report was accompanied by an elaborate history of the several religious communities.
The enumerations in Ireland exhibit statistical novelties of a totally different kind. In 1841 it was resolved, as that country so totally depended on the amount of its agricultural produce, to obtain the statistics of its rural economy. The surface of the country was divided unde v five heads—arable, plantations, uncultivated, towns, and water; and, with a view to these divisions, a return was made of the character of each farm or other agricultural allotment, with the quantity of live stock and other relevant facts. The attempt was found so successful, that it was renewed with more full effect in 1851, producing 727 tabular folio pages of very valuable information.
Census of 1861.
The enumeration in 1861 was the seventh census of England, and was taken under the superintendence of the Registrar-General, under the powers conferred by three Acts of Parliament, applicable respectively to England, Scotland, and Ireland. By this census it was found that the population of the United Kingdom was 29,321,288, and that of each pf the four divisions of the kingdom was as follows :— England and Wales, 20,228,497; Scotland, 3,096,808; Ireland 5,850,309; islands in the British seas, 145,674.
The population of England and Wales, excluding the portion of the army, navy, and merchant seamen abroad, was 20,066,224. The annual rate of increase per cent, of the population in the United Kingdom, in the ten years 1851-61 was 0 55, viz.:—England and Wales, 144; Scotland, 0’58; Ireland, (decrease), 123; and islands in the British seas, 0O2.
The local machinery by means of which the census of England was taken in 1861 differed in no material respect from that employed in 1851. In subdividing the country the registrars were to see that the enumeration districts were not too extensive or too populous; they were not to contain more than 200 houses in towns, so that they could be enumerated by an active man within the compass of a single day. An important feature consisted of the publication of the number of houses and population in “civil counties” and in lieutenancy subdivisions. The number of blind and of deaf and dumb from birth, the number of foreigners, and the number of naturalized British subjects, were shown in 1861 for the first time. In the case of Scotland, the enumeration was for the first time required to be made apart from that of England, under the superintendence of the Registrar-General of Scotland. From 1855 Scotland had the advantage of a national system of registration, and the census of 1861 was therefore no longer taken through the agency of the sheriffs in counties, and of the provosts or chief magistrates in the royal and parliamentary burghs, but through the agency of 1001 local registrars of births, deaths, and marriages, assisted by 8075 enumerators. The forms and instructions issued were all based on, and in most cases virtually the same as, those used in England in the census of 1851. “Flats ” were not reckoned as houses in Scotland in 1861. All the details as to population, <fcc., in 1851, were for civil parishes and civil counties only. In 1861 the returns were for registration districts and registration counties. The novel features in 1861 for Scotland, which merit special notice, were the number of families in every parish and county, the number of houses having windows, and the number of children between the ages of five and fifteen years in attendance at school. In 1861 it was ascertained that the islands in Scotland amounted to 787 in number, and of these 186 were inhabited by one or more persons on the census day.
The tables in 1861 relating to the ages, conjugal conditions, birthplaces, blind, deaf and dumb, and to the occupations combined with ages, were published uniformly with those in 1851. A comparison can be made of the occupations at the different ages with the death registers, and the comparative mortality prevailing among the different classes and ranks of society can now be ascertained as well as the comparative healthiness of each occupation or trade.
Ireland, 1861.—The enumeration of the people of Ireland in 1861 was effected for the first time altogether (excepting the metropolis) by the officers and men of the constabulary force. The returns show—for provinces, counties, baronies, parishes, towns, and parliamentary boroughs—the area, population, and number of houses, the number of families with their pursuits and the means upon which they were dependent, also the religious professions and education of the inhabitants, the number of families living in each house, and the amount of their accommodation by counties in rural and civic districts, the birthplaces of the people, their condition as to marriage, and their occupations for each county, also tabular summaries relating to vital statistics.
The ages of the people in Ireland were incorrectly returned in 1861, and in the tables of the occupations of the people a large number are left unclassified.
Iu the Channel Islands and the Isle of Man, the census was taken in conformity with instructions from the Home Office, by the respective Lieutenant-Governors acting in conjunction with the English central office.
First Imperial Census, 1871.
In 1871 the first Imperial census of the British empire was taken, and the population was found to amount to 234,762,593, living upon 7,769,449 square miles of territory; comprising England and Wales, with a population of 22,856,164; Scotland, 3,392,559; Ireland, 5,449,186: islands in the British seas, 147,470; and British Colonies and Possessions (exclusive of the army, navy, and merchant seamen abroad, as they are included in the populations just quoted), 202,917,214. The population of England and Wales, excluding the portion of the army, navy, and merchant seamen abroad, was 22,712,266. The annual rate of increase per cent, of the population in the United Kingdom in the ten years 1861-71 was 0 - 83, viz. :— in England and Wales, 123; Scotland, 092; Ireland (decrease), 0 - 71; islands in the British seas, 0 , 12.
In this great undertaking the work was subdivided for convenience of execution. In conformity with the Act of Parliament the census of England and Wales was taken by the Registrar-General, assisted by Dr Farr and J. T. Hammick. The population was enumerated in one day (3d April 1871) by 32,543 enumerators, employed under 2195 registrars and 626 superintendent registrars.
Instructions were issued to the superintendent registrars, and to the registrars of births and deaths, for their guid ∙ ance in taking the census, and one of the duties requiring their earliest attention was the division of the country into enumeration districts, and just here the registrar had an important duty to fulfil, for he was directed to see that the boundaries of parishes, townships, ecclesiastical districts, municipal and parliamentary cities and boroughs, urban sanitary districts, and all other of the various conflicting and confusing territorial subdivisions of the country, were duly indicated. Most of the registrars laid down the boundaries of their enumeration districts on maps, so as to ensure that no part of their districts was omitted.
These plans of division were first submitted to the various superintendent registrars, and finally to the Registrar-General for approval. The 32,543 enumerators were required to be intelligent, trustworthy, and active, to write well, and to have some knowledge of arithmetic. They were not to be infirm or of weak health, not younger than 18 years, nor older than 65; and they were required to be persons likely to conduct themselves with strict propriety and civility. 627 of the public or charitable institutions were enumerated by the masters or chief resident officers.
In order to facilitate the proper designation of all roads, streets, and houses, a circular was addressed to the mayor of each municipal city and borough, and to the chairman of every local board, suggesting that previous to the census all unnamed roads and streets should be named, and all houses numbered. Circulars were also addressed to the press explaining the objects and uses of the census, so as to aid in securing complete and correct returns,
The enumeration of 1871 was a “nominal” census.
The householders’ schedules were delivered by the enumerators prior to 3d April. Every separate occupier received a schedule, and this schedule was so prepared in accordance with the 4th section of the Census Act, that, as shown in the annexed example, the name, sex, age, rank, profession or occupation, conjugal condition, relation to head of family, and birthplace of every person who abode in any house, on the night of Sunday, 2d April 1871, might be returned; the blind, deaf and dumb, imbecile, or lunatic were distinguished. At the census of 1871 the numbers of lunatics out of asylums and of imbeciles or idiots were ascertained for the first time.
[table]
The despatch of forms and instructions began as soon as the office in Craig’s Court was opened, and included 6,500,000 householders’ schedules, weighing about 41 tons. The enumeration books and forms sent out weighed in the aggregate about 54 tons. But in addition to these there were sent out from the central office no less than 115 different descriptions of printed instructions, forms, and circulars, including a calendar, so that certain operations might be summarized.
The metropolitan, the municipal, and the county police afforded help in enumerating the houseless population. The strength of the navy, in forms prepared for the purpose, was returned by the Admiralty; the merchant seamen, in port or out at sea, by Her Majesty’s Customs, and by the Registrar-General of merchant seamen; and His Royal Highness the Field-Marshal Commanding-in-chief supplied full returns of the various particulars respecting the army. The numbers of British subjects in foreign states and in India and the population of the Colonies were obtained through Her Majesty’s Secretaries of State for Foreign Affairs, India, and the Colonies.
Welsh schedules were issued for the use of a certain number of the Welsh people, and it was found that in 17,276 of these the particulars were inserted in the Welsh language, and had to be translated.
The enumerators were instructed to consider a house as comprising all the space within the external and party walls of a building, whether occupied by only one family, or by several families living in distinct stories or apartments; and they were also instructed, when delivering their schedules, to make a record, in a “memorandum book” supplied for the purpose, of the road or street in which every house was situated, and of the name or number of each house, and to state the number of schedules left at each house. Uninhabited houses, houses building, churches, chapels, and all other public buildings were likewise noticed in this book. With the assistance of the police, the enumerator had to make or procure returns of all persons not dwelling in houses, but sleeping in barns, sheds, caravans, or tents, or in the open air.
Special schedules were printed for the enumeration of persons in public institutions, on board vessels, or in charge of boats and barges employed in inland navigation.
Persons travelling during the night of Sunday, 2d April, were to be included in the schedule of the proprietor or manager of the hotel or inn, or in the schedule of the occupier of the house, at which they arrived on the morning of Monday, 3d April. Persons engaged in work away from home during the night of Sunday, 2d April, were to be included in schedules left at their houses.
The cause of an unusual number of persons being temporarily absent or present was to be noted.
VVhen the enumerators had collected and arranged their schedules, they were copied into enumeration books provided for this purpose, and a summary was made of the population of each civil parish, township, or place, and these books were delivered to the registrar on or before 10th April, by whom they were examined and corrected, and forwarded on or before 24th April to the superintendent registrar.
The enumeration books were then revised by the superintendent registrar, who was at this stage required to fill up a printed form, showing a summary of the houses and population in each sub-district, municipal or parliamentary city or borough, <fcc., and transmit the whole to the Census Office in Craig’s Court.
On receipt of the books and summaries by the RegistrarGeneral, the abstract required by the 9th section of the Census Act was prepared and laid before Parliament on 20th June 1871.
As soon as the preliminary report, consisting of 137 pages, was published, the work commenced at the central office of revising the enumeration books so as to ascertain the correct number of houses and population in each of the various subdivisions of the country, and in order to ensure accuracy in the preparation of this work, statements of the houses and population of ecclesiastical districts and other local subdivisions were submitted to the clergy and to the various local authorities for approval. The registrars were required to state the cause of any remarkable increase or decrease of population, and a proof of the tables relating to each county was submitted to the clerk of the peace for correction and approval.
The abstracts relating to the occupations and ages of the people were a great work. It was considered desirable not only to take out the number of persons of each sex in each occupation, but the number at each quinquennial period of age; for without this information the relative salubrity of the professions, and a great variety of other important questions could not be determined.
In this arrangement of the people according to occupations and ages, they had not only to be classified in different orders, but it was necessary to find a place for every one of the 23 millions of population, so as to be enabled to pass them rapidly and distinctly in review.
The classification of the living according to their professions. and occupations in combination with age has opened up a new field of inquiry, and the question can now be determined for sanitary purposes, and for life insurance, what effects the different professions and occupations have on health, at different ages. Formerly, the mean age at death was taken to show the healthiness or insalubrity of certain occupations. But the mean age at death depends upon many circumstances besides health, and among others, upon the ages of the living, which vary in proportions in almost every profession, according as it is a profession that people enter early or later in life. Insurance offices and friendly societies will find the results of the third volume of the Census of England in 1871—in conjunction with the 14th Annual Report, and the two Supplements to the 25th and 35th Annual Reports of the Registrar-General, recording the mortality of persons at different occupations—of great use to them in their trans actions, for it is evident that the lives of farmers, for example, may be safely insured at much lower rates than the lives of licensed victuallers.
Many important results for “civil counties” were published in 1871 that will not be found in the census returns of 1861, such as a description of the territorial, executive, and ecclesiastical divisions in each county, more exact areas from the ordnance survey, additional columns showing the houses under three headings, and distinguishing the sexes of the population, additional columns showing area, number of electors, and of members within the parliamentary limits, including and excluding represented boroughs, petty sessional divisions, wards of municipal boroughs, comparative tables of houses and population in 1861 and 1871, tables showing the number of families in each civil parish and township, and results showing the residue of area and population in mother parishes.
Also under “registration counties” many new particulars were published in 1871, such as maps of counties and districts, showing the population of large towns, and the particulars as to reformatory and industrial schools, and as to Her Majesty’s ships.
The census of Ireland in 1871 was taken by the Registrar-General of Ireland, assisted by two other commissioners, W. R. Wilde and G. W. Abraham. The enumeration (except in the metropolis, in which 178 of the Dublin Metropolitan Police discharged that duty) was effected by 4536 members of the Royal Irish Constabulary, whose local knowledge and previous experience as enumerators on three former occasions rendered them peculiarly well suited for this undertaking. The country was divided into 248 districts, each under the charge of a sub-inspector. Printed instructions were furnished, and, when necessary, the enumerators were supplied with maps, conveyances, boats, and in the Irish-speaking districts with interpreters also. In conformity with the Census Act an abstract of the census of Ireland was published on 14th June 1871, showing the number of houses, families, and population in each county and city, and in certain corporate towns in Ireland. The numbers in the different religious professions were also returned as directed by the Act. The revised and final numbers, for Ireland record the population in 1871 at 5,412,377, showing a decrease since 1861 of 386,590 or 6*67 per cent. In the previous decenniad, 1851-61, the population had decreased by 753,418, or 11 ’ 50 per cent. The decrease in the population of 1871 is stated to be chiefly attributable to emigration. From 1st May 1851 to 31st March 1871 the loss by emigration amounted to 2,024,609. Results relating to the ages, civil conditions, occupations, birthplaces, religion and education of the people are published in the census returns for Ireland; and instructions were issued by the Government, directing that the results of the census of 1871 for each of the three divisions of the kingdom should be published, with a view to uniformity. Some of the principal results have thus been assimilated in 1871, but there is still a considerable want of uniformity, especially as regards, the ages of children in Ireland. A volume devoted entirely to vital statistics is not the least interesting of the results of the census of Ireland in 1871.
As regards Scotland, the results of the census of 1871 are nearly uniform with those of England. In addition to the ordinary particulars, inquiries were made as to the number of children from five to thirteen years of age receiving education, and the number of rooms with windows. This census was taken by the Registrar-General of Scotland and Dr Stark, through the agency of 1016 local registrars, assisted by 8342 enumerators, whose appointments were approved by the sheriffs in counties, and by the chief magistrates in boroughs. The returns for Scotland in 1871 contain some new features, such as the ecclesiastical subdivisions of civil counties, grouping the population of each civil county into three great divisions, according as they inhabited towns, villages, or rural parts; the particulars as to houses, families, and population relative to parliamentary constituencies of counties are also new. As there were no registration districts in 1851, no attempt was made in 1861 to show the relative increase or decrease of the population from 1851. But in the census of 1871, this want was supplied by giving the number of the population in each district for 1861, for comparison with that of 1871. The number of children from five to thirteen years of age in the receipt of education also appears for the first time in the returns for 1871.
The census of the Island of Man and of the Channel Islands was taken by the lieutenant-governors according to directions issued by the Secretary of State for the Home Department, but the forms and instructions were issued by the Registrar-General of England. The population of these islands on the aggregate in 1871 was 144,638, viz.:—Isle of Man 54,042, Jersey 56,627, and Guernsey and adjacent islands 33,969. No census of the Isle of Man was taken earlier than that of 1821, when 40,081 persons were enumerated. Enumerations of the inhabitants of Jersey were made in 1806 and in 1815, in which years the population was respectively returned as 22,855 and 22,763. The first census of Guernsey, <fcc., was taken in 1821, and was 20,302, The population of these islands has remained nearly stationary at the last three censuses.
The cost of the census, incurred at the central office, and in payment of local officers in each of the three countries in 1871, was £5, 5s. 7fd. per 1000 of the population for England and Wales, £8, Is. 4d. for Scotland, and £7, 2s. 7d. for Ireland.
The results of the census of 1871 are multitudinous, and it is impossible here to enumerate all the facts of interest which the returns disclose. Those relating to parliamentary constituencies show that of 95 counties, or divisions of counties, the population had increased since 1861 in 82, while in 13 the population had decreased. Out of 198 boroughs which returned members to Parliament the population had increased in 156, while in 31 the population had decreased. In the year 1868 the boundaries of 68 boroughs were altered, and in some of these the increase of population may be due to enlarged area. Eleven boroughs had been formed since 1861. Some of the elaborate results afford the means of solving important questions relating to the industrial organization of the country, and show to what extent there has been a diminution in the numbers employed in the cultivation of the soil, and to what extent there has been a consolidation of farms, affecting the average size of the holdings. Other results show the ages of the married and unmarried population, and these, combined with the returns of the ages at marriage, furnish the means of determining very important questions, such as the probable duration of the joint lives of husband and wife, the annual rate of marriage at different ages, and so on. Tables showing the number of foreigners resident in England and Wales, and the number of British residents abroad, throw much light on many points relating to the social condition of the people. Not the least useful of the results is an interpolated table of the population at each year of age, furnishing school-boards with the means of ascertaining the proportional number of children at the school ages, and enabling life insurance offices and scientific and other authorities to solve many important problems. (w. f.) Census of the United States.
The census in the United States is a political necessity. To this, and not to the general appreciation of the claims of statistical science is due the fact that the systematic periodical enumeration of inhabitants was there undertaken earlier even than in England.
Had the government set up in 1787 been a pure confederation, like that which preceded it, into which the States entered as equal bodies, the census would not have been a condition of its existence; while rude estimates founded on the number of the natural militia, the arms-bearing population, or on the number of horses, or on the rçcords of births and deaths, might long have answered all administrative purposes. On the other hand, had the Constitution of 1787 erected a single, self-sufficient government, a simple sovereignty, the census need not have been provided for in that instrument. Representative power might have been apportioned approximately according to common fame, or numbers might have been rejected as the measure of political power, as in England, and even as in Connecticut, alone of all the States of the American Union, to-day.
But the mixed form of government established by that constitution, the only form of government which was then possible, by which the pre-existing States maintained their right to exist and to act for themselves in all strictly internal affairs, while for all national purposes political power was to be exercised by a double rule, partly through the States acting as equal bodies, and partly according to population irrespective of State lines, positively required, not as a means of administrative efficiency, but as an essential condition of its own existence, that the inhabitants of the United States should be periodically enumerated. Accordingly, we find in the Constitution of 1787a provision for a census to be taken every ten years, the first enumeration to be made “within three years after the first meeting of the Congress of the United States.”
It would appear from this brief review that the eulogium of Moreau de Jonnès was scarcely merited, when he declared that the United States present a phenomenon without parallel in history,—“that of a people who instituted the statistics of their country on the very day when they founded their government, and who regulated by the same instrument the census of inhabitants, their civil and political rights, and the destinies of the nation.” As a matter of undoubted history, the provision for a national census was incorporated in the constitution solely from political, not at all from philosophical, considerations. Pending the actual accomplishment of the first enumeration, an estimate of the population of the several States was adopted as a provisional basis of representation in Congress.
The first census was taken in 1790, the returns all being referred to the 1st of August of that year, although the work of canvassing was extended over a considerable period. As was the case with the first British census eleven years later, and as is likely to be the case with the first census of any community, great disappointment was felt at the result, and dissatisfaction at the methods of enumeration was loudly expressed. Mr Jefferson, then Secretary of State, was careful in sending abroad copies of the published tables, to impress it on the minds of his correspondents that the returns fell far short of the truth, and even went so far as to supply the omissions which he assumed to exist. The results of later censuses, however, have established the substantial accuracy of the first enumeration, and shown that the dissatisfaction felt at the time was but the inevitable disappointment of over-strained anticipations. “To count,” says Dr Johnson, “is a modern practice; the ancient method was to guess; and where numbers are guessed, they are always magnified.” General censuses were taken in 1800, 1810, and 1820, with exact intervals of ten years. In 1830 the date of enumeration was fixed on the 1st of June. This change made the interval between the fourth and fifth censuses nine years and ten months only, reducing the nominal ratio of increase between 1820 and 1830, by about two-thirds of 1 per cent. The succession of census was then resumed with exact intervals of ten years. The ninth census was taken as of date June 1, 1870.
Peculiarities of the Constitution of the United States have been spoken of which rendered the census a political necessity, and secured its adoption there earlier than in countries far more advanced in statistical science and in the arts of administration. To the same peculiarities are attributable the features by which the census of the United States has come to differ from the census of other countries. The reservation by the States of all rights not granted to the general Government makes it fairly a matter of question whether purely statistical inquiries, other than for the single purpose of apportioning representation, could be initiated by any other authority than that of the States themselves. That large party which advocates a strict and jealous construction of the constitution would certainly oppose any independent legislation by the national Congress for providing a registration of births, marriages, and deaths, or for obtaining social and industrial statistics, whether for the satisfaction of the publicist, or for the guidance of the legislature. Even though the supreme court should decide such legislation to be within the grant of powers to the general Government, the distrust and opposition, on constitutional grounds, of so large a portion of the people, could not but go far to defeat the object sought.
In this political difficulty, the unquestionable provision of the constitution for a decennial census has been taken advantage of by all parties to secure much statistical information, which is not usually, and perhaps is not properly, connected with a census. Nor can there be any doubt that the introduction of new schedules of inquiries has, since 1850, somewhat impaired the efficiency of the census in its original constitutional function of making a count of the inhabitants of the several parts of the country for the purpose of distributing representation in Congress. As the census has widened, it has weakened. More has been put upon the enumerator than he could well carry. His attention has been distracted by the multiplicity of objects presented; the great number of inquiries has also perplexed and irritated the body of citizens; while the result of protracting the canvass to obtain additional information has, in the incessant changes of population, especially in cities and in manufacturing villages, allowed not a few to escape enumeration altogether.
The first enlargement of the scope of the census was in 1810, when the agents of the census were by law required to take an account of the several manufacturing establishments and manufactories within their several districts. The same requirement was contained in the Act for the census of 1820; but the results proved of so little value that the attempt was altogether abandoned in 1830. In 1840 a manufacturers’ schedule was used, and inquiries respecting schools were incorporated. It was at the seventh census, 1850, however, that the main enlargement was effected. By the Act of May 23 of that year, the census was extended to cover the several subjects of mortality, agriculture, productive industry, schools, churches, newspapers, <fcc., <fcc.
The agencies and methods of enumeration in the census of the United States have been but little changed from the first, and are to-day antiquated and ineffective in a high degree. By the Act of 1790, it was made the duty of the marshals of the United States courts in the several judicial districts to take the enumeration, appointing therefor as many assistants as they should deem necessary. The enumeration was to commence August 1, and to close within nine calendar months thereafter; the returns were to be filed with the clerks of the several courts, while the marshals were to forward to the president “the aggregate amount of each description of persons within their respective districts.” In 1800 the supervision of the census was placed with the Secretary of State. The machinery oι enumeration remained until 1850 the same as in 1790.
In 1810 the agents of the census were placed under the direction of the Secretary of the Treasury as to the enumeration of manufactures required; but in 1820 the entire charge remained with the Secretary of State. In 1850, the supervision of the census was committed to the newly created Department of the Interior, and the period allowed for enumeration was restricted to five months. As a matter of fact, no enumeration has been completed within that time. The United States marshals and their assistants remained the agents of the census office; and a superintendent of census was appointed. In preparation for the census of 1870, a committee of the House of Representatives, under the chairmanship of General James A. Garfield of Ohio, presented a bill which sought to bring the census as nearly up to the demands of modern statistical science as the peculiar conditions of settlement and occupation in the United States would allow. The report of the committee dwelt strongly on the essential viciousness of a protracted enumeration, and on the general unfitness of the marshals of the courts to perform the duty. The bill, however, did not propose a single-day enumeration, but allowed a month for its completion. The schedules of inquiries were largely increased in recognition of the impossibility of obtaining statistical information under authority of the United States Government, except under cover of the constitutional provision for a decennial census. The rank and power of the superintendent of census were increased.
The bill of the committee was carried through the House of Representatives by large majorities, but failed in the Senate. The census of 1870 was consequently taken under the Act of 1850 with two or three inconsiderable amendments.
The cost of each census has been as follows :— [table]
In the census of 1870 the work of enumeration was performed by 6572 assistant marshals, supervised by 61 marshals of United States courts.
Prior schedules so called, that is, schedules to be left at the residence in advance of the enumeration, to be filled up by the head of the family, have never been used in a census of the United States, though adopted in nearly every European country. (f. a. w.)
