REPORT OF THE SUPERINTENDENT OF THE CENSUS. NEW HAVEN, CONN., SIR: In compliance with your instructions, I have the honor to submit the following suggestions respecting the census of the United States required, by the first article and second section of the Constitution, to be taken in 1880. By the twenty-third section of the act of May 23, 1850, it is provided "that if no other law be passed providing for the eighth or any subsequent census of the United States, on or before the 1st day of January of any year when, by the Constitution of the United States, any future enumeration of the inhabitants thereof is required to be taken, such census shall in all things be taken and completed according to the provisions of this act." It appears, thus, that legislative sanction and provision already exist for the taking of the tenth or any subsequent census of the United States. I cannot, however, refrain from the strongest assertion of the impolicy of continuing to enumerate the population of the country under the act of 1850. That law was passed when statistical science was in its infancy and the art of collecting statistical data was yet in a rude and imperfect state. The scheme of enumeration then erected was never a good one, while, so far as it was suited to the then existing conditions of settlement and industry, the country has been steadily growing away from it through the period that has since elapsed. To take the tenth census under the provisions of the act of 1850 would be to secure the minimum of statistical result at almost a maximum of annoyance, delay, and expense. The principal points at which changes in the census scheme as at present existing seem to be required are the following: 1st. As to the persons who shall superintend the census in the severat States. The act of 1850 provides that the marshals of the several judicial districts of the United States shall direct and superintend the enumeration, with the power of appointing assistants actually to conduct the work. The object of this provision would seem to have been to charge this duty upon some body of officers already in service, in order to save either the trouble or the expense of bringing a new set of officials into existence for the purpose. But the work of census-taking is so far exceptional and unique in its requirements as the rather to create the presumption that those who are to undertake it should be selected for the purpose. Between the duties of the United States marshal and those of a supervisor of the census there is no very clear or close connection. A man might be the most energetic and efficient marshal on the lists of the Department of Justice, yet have less than ordinary qualifications for organizing and superintending the taking of the census-a duty demanding a high degree of clerical capacity and a fitness clearly to conceive and forcibly to impart to the subordinate enumerators the many precise and delicate distinctions which are required properly to answer inquiries of the census schedules. Other objections, however, withstand the delegation of United States marshals to this duty. One is, that the formation of judicial districts is determined by reasons altogether different from those which should determine the formation of census grand divisions. Southern Florida, with its innumerable reefs and keys, is periodically strewn with wrecks. From the same facts, taken in connection with its proximity to the islands of the Gulf, it affords great facilities for smuggling. For these reasons it is most properly constituted a judicial district, and for that reason most improperly it is, by the law of 1850, constituted a census district, with 5,775 inhabitants enumerated by a single assistant marshal, with a United States marshal to overlook the important operation. Northern New York, with nearly two and a half millions, also constitutes a census district, and its six or seven hundred assistant marshals are all to be instructed and overlooked by one United States marshal. Delaware is a district; so is Massachusetts. Idaho is a district; so is Indiana. It is hardly necessary to say that, if superintendence is of any account in census work, the superintendence which is provided by the law of 1850 must be of the least account possible. [Report of the Superintendent of Census, November 21, 1871, p. xxiv.] But, again, the act of 1850 charges the duty of supervising the census upon a body of officers most of whom have already as much labor and responsibility as they can in justice to themselves or to the government manage to get along with. In the largest districts, especially those which have great manufacturing interests and contain many cities, where the liability to error rises to its maximum and the need of superintendence is most felt, the marshals can give to the census only a divided attention and a fragment of their time. Under these circumstances, the best thing the marshal can do is to entrust the whole census work thus brought into his office to a deputy, who may not even be known by name at the census office. The marshal holds himself amply excused, by the number and conflicting nature of the duties imposed on him, from giving his personal attention to the service. The deputy, who may be well chosen or ill chosen for the purpose, does the work anonymously and without any appreciable degree of official responsibility, engendering thus the worst vice of public administration. But while I would, for these reasons, strongly urge another rule of appointment for those persons who are to take the census of 1880, I would recommend that United States marshals be made eligible, with the consent of the Department of Justice, to that office. It might be found that, in a dozen or twenty districts, as they are at present formed, the marshal might be the person who could most advantageously conduct the enumeration, reference being had to his personal qualifications for the work, his acquaintance with the population of the district, the size of the district and its conditions of settlement, and the nature and extent of the duties already demanding the attention of the marshal. The bill providing for the taking of the ninth census, which passed the House of Representatives in 1870, contemplated the appointment of one district superintendent of census for each Congressional district. Such a provision would be preferable to that now existing; but, in my opinion, a better superintendence of the work of enumeration could be furnished at a lower cost. If the Department of the Interior were authorized to appoint as supervisors of census one or more persons in each State, the aggregate number of such officers not to exceed one hundred and fifty, it being left to the department to arrange the several su pervisors' districts to suit the requirements of enumeration, a better result would be obtained than that contemplated by the bill of 1870, at not more than half the cost. The difference, however, would not merely be a saving in cost. Congressional districts are made up to meet a single requirement, namely, the division of the State into districts of approximately equal population. No distinction is made of city and country, of agricultural, manufacturing, commercial, and mining populations, of fertile valleys easily traversed and fully settled, and of mountain regions difficult of access and sparsely inhabited. It may be said without any suspicion of exaggeration that one hundred thousand persons, under one set of circumstances as to occupation and location, may require more superintendence for census purposes than half a million under other circumstances. Indeed, no city, however large, should have more than one responsible head for its enumeration. So far the bill of 1870, already referred to, accorded with the present recommendation, it being left, by that bill, with the Secretary of the Interior to arrange for the superintendence of the census in large cities without regard to the lines of Congressional districts. The number of census supervisors being thus fixed, and the apportionment of the number being left to the discretion of the department, it might be provided that none of these officers should receive compensation for a longer term than days, or at a higher rate than $— per day. 2d. As to the appointment of the actual enumerators. Under the act of 1850, the power of appointing the assistant marshals who were charged with the actual enumeration was vested in the United States marshal absolutely, the department at Washington having no power to reject for any cause. It would seem that good administration would require that the department of the government intrusted with the census should have the same power over the actual enumerators which the Treasury Department or the Post-Office Department has over appointees of its local offices. Subject, therefore, to confirmation or rejection by the Department of the Interior, it is respectfully recommended that the appointment of enumerators be intrusted to the census supervisors in their several districts. 3d. As to the formation of census subdivisions. The same reasons which require that the appointment of enumerators should be subject to approval or rejection at the central office would seem to make it necessary that the formation of the enumeration districts should be submitted for a like revision; but in the latter case there is peculiar occasion for such a check upon the action of the local authorities, to prevent the creation of subdivisions too extensive or too populous to be enumerated within the time allowed by law. Much difficulty was experienced from this cause in 1870, several marshals insisting, against the advice of the Census Office, on assigning to assistant marshals districts which could not possibly be canvassed in compliance with law in the prescribed time, the result being either the undue protracting of the enumeration, or else the illegal letting out of the work to unauthorized parties. 4th. As to the compensation of enumerators. This is doubtless the most important and most difficult question to be raised in providing for the taking of a census. In the United States it is of peculiar difficulty, owing to the vast range which exists in the conditions of settlement and occupation. It is, in my judgment, impracticable to frame a single rule which, with advantage to the government and with justice to enumerators, can be applied alike to crowded cities, to ordinary agricultural communities, to the population of mountainous regions, to prairie settlements in the newer States, and to the scattered inhabitants of the grazing or mining Territories. The act of May 23, 1850, did indeed attempt to set up a rule which should meet all conditions of settlement. It provided that the assistant marshal should receive two cents for each person enumerated and "ten cents a mile for necessary travel, to be ascertained by multiplying the square root of the number of dwelling-houses in the division by the square root of the number of square miles in each division, and the product shall be taken as the number of miles traveled for all purposes in taking this census." For this rule it was claimed that its action was compensatory in the degree needed to secure substantial equity. In the closely settled regions, it was urged, the enumerator would receive chiefly a per capita allowance; the district being small, the mileage would be insignificant. In sparsely populated regions, on the other hand, the enumerator would obtain but a small portion of his compensation in the form of a per capita allowance, while the mileage would constitute his real remuneration. It cannot be denied that there is in this rule a certain tendency toward equalizing the compensation of enumerators; but a careful study of the workings of the system at the census of 1870 has satisfied me that it allows great injustice to be done as between enumerators laboring with equal energy and zeal, while costing the government in the aggregate far more than would be needed to secure quite as thorough an enumeration under a rule which permits the exercise of administrative discretion, within certain limits, in adapting the rates of compensation to the varying conditions of settlement and occupation. There were hundreds of enumerators in 1870 who earned six and eight dollars a day quite as easily as hundreds of others earned two and three dollars. In exceptional cases, the disproportion of compensation was even greater. The general result was unfairness as toward enumerators and unnecessary expense to the government. I would therefore suggest that at the census of 1880 an aggregate amount be appropriated by Congress for the compensation of enumerators, to be applied by the department, in its discretion, subject to the provision that no person so employed shall receive more than dollars a day for each full day of hours, or in proportion for any fraction of a day, the enumerator's statement of time occupied in his work being verified under oath. Such a system would entail upon the department a great responsibility and no little labor; but that responsibility should be borne by some one, and that labor performed, in justice both to the government and to the body of enumerators. The soundest system, administered with the highest discretion, will not avoid injustice to individuals, but the department could not fail, in its use of such a lump appropriation, to reach a far more equitable apportionment of pay to work than the operation of the rule of 1850 permits; while, by saving the wasteful allowances in many cases resulting from the application of the present rule, the whole enumeration could be accomplished at an expense of not more than 80 per cent. 5th. As to the time to be occupied in enumeration. In England, or in some other countries of Europe, the census is taken as nearly instantaneously as possible, and the people of the kingdom are, as it were, photographed in the position they occupied at a given moment. To be more specific, the enumeration takes place with reference to a single night, and, schedules having been distributed in advance, |