PAPERS AND DISCUSSIONS.

SCHOOL STATISTICS AS THE BASIS OF LEGISLATIVE OR OFFICIAL ACTION-WHAT SHOULD BE COLLECTED, AND HOW?

HARVEY M. LA FOLLETTE, STATE SUPERINTENDENT OF INDIANA.

All legislative or official action having for its purpose the improvement of our public-school systems must, if wise, be based upon experience and the past observation and study of actual conditions. These conditions and results can only be definitely and analytically canvassed and studied through the medium of reliable statistics. Facts that are counted and classified form the natural, though numerical, basis for the study and interpretation of educational growth and progress, and, with the essential explanatory and descriptive text, make up the only rational guide for such action.

The value of such statistics is determined by their accuracy, the practical purpose kept in view in their collection and arrangement, and their universality or community of origin, forms and terms.

The careless collecting of so-called "statistics" that are largely estimates, and the irregular recording and reporting of them to central authorities, can only serve to supply false bases, unwarranted conclusions, and, by consequence, injurious theories and acts. Each class of statistics must be collected, recorded and reported for some specific purpose, and such aim be kept clearly in view by every one concerned, to be of the greatest value. If statistics of attendance, for example, be recorded not in their true significance as the exposition of the actual return made for means expended, and as the reliable suggestion of the movement and growth of general education—but rather as a neighborhood comparison between adjoining schools and communities, and a measure of the personal interest of teachers and pupils—the real purpose will be lost sight of and the statistics be so warped as to be of little value.

The more universal statistics are, the more valuable they become. If it were possible to secure, not only a national but an international basis of educational statistics, it would be the greatest possible step toward a rational and universal basis for wise school legislation, and would afford to every central authority the opportunity to profit by every improvement in organization, method, and administration; but it is an error to substitute generality for universality, and so to kill the definite character and immediate or even local application of educational statistics in the attempt to render them universal. While the possibilities of universal school statistics are great, yet

in the United States there must be, for long years to come, at least, as wide differences in many facts and figures to be collected and classified, as there are different State constitutions and different policies and theories back of them.

To secure the highest and best good from school statistics as a basis of legislative or official action, as well as from a general historical and sociological standpoint, the statistics of education must be taken as complementary to other social statistics, as those of illiteracy, crime, morals, wealth, etc., and all these factors be duly considered in determining the course and purpose of such action. The determination of the educational conditions and needs of the State in the aggregate is an absolute essential to wise legislation. By these conditions and needs the law-makers should be guided in determining the character of general organization of school systems, the limitations of local powers, the establishment and maintenance of library systems, the regulation and requirement of attendance, the basis of apportionment of State school revenues, as well as many other important conditions of social and of criminal legislation. The proper study of such facts and their adoption as a basis for legislation have invariably resulted in the complete divorce of school organization from other systems of civil organization, and the establishment of a system controlled by school officers whose duties do not extend beyond the discharge of the functions of the schools. The value of such separation of the school interests from other local interests, such as highways, bridges, and ditches, the care of the poor, etc., is of itself a very great gain. It has resulted, too, in the greatest possible simplicity of organization, which, while observing all proper limitations of local powers in the levying of taxes and similar functions, yet places the powers attendant upon such administration in the hands of a single individual, where the weight of responsibility may also fall, and thus secures an intelligent, responsible, and directive administration. It has secured and must ever secure a realization of the fact that for a commonwealth to expend several millions of dollars annually for the education of her school children and yet fail to require their attendance upon such or similar schools, is an act of folly, both educationally and commercially, and without rational justification.

Perhaps the failure of many of our States to appreciate the lessons of their statistics in this particular, and the indifference in other commonwealths as to the strict enforcement of attendance laws already in existence, are largely explained by the fact that we have drifted into methods of keeping school records and reporting the alleged conclusions therefrom to school authorities, that are based, not upon a basis of perfect attendance, but upon membership only. In other words, in the great majority of States the school statistics are based, not upon the actual attendance based upon the number of pupils enrolled, each pupil being charged with attendance for the full number of days that such school is open, but only for the time that such pupil is a "member" thereof. All the time lost by pupils who should be in the public schools from

the beginning of the school term, but who do not enter until the term is far advanced, in many cases until one-half thereof is gone, is not included in such reports of time lost; and in the great majority of cases, perhaps, the time lost by pupils after three days of successive absence is not recorded or included in the school registers and reports of attendance, but after such an absence of three days such pupils are dropped from the rolls and are no longer recorded as members of the school. And further, while perhaps in the majority of schools a pupil who has lost his membership, on returning to school is only "reinstated" and the accounting of his attendance resumed from that time forward; in a vast number of schools such a pupil upon his return is enrolled as a new pupil, and so the per cent. of the enrollment upon the enumeration as returned in the annual reports is greatly increased. Moreover, the methods of keeping school records that have long obtained in the public schools, based upon the percentage system, lead in many cases to annual summaries that are very erroneous and misleading. Thus, for example, in a school of six months there might be enrolled for the first month twenty-four pupils, the average daily attendance therein, as based upon membership, showing an average of eighteen, or, as recorded at the close of the month, 75 per cent.; for the second month, an enrollment of thirty-five and an average attendance upon the membership of twenty-eight, or 80 per cent.; an enrollment for the third month of forty and an average attendance of twenty-eight, or 70 per cent. If in the following three months the ratio of membership decrease in like proportion, we will have as a final summary and average attendance for school term, by months, 75 per cent. Now it is probable that in such school there may have been enrolled within such school term forty-eight different pupils. In the annual report of the teacher to the trustee there is included a complete list of such pupils by name and their class-standing in the matter of scholarship. In making his verified annual report of attendance within such school corporation, the trustee includes from such school, as the average attendance, 75 per cent. of the total enrollment, namely, 48, or 36 pupils as the actual number that have been in daily attendance upon such school throughout the term of six months. Now as a matter of fact the average attendance as based upon membership in such school has been but 24; and if, instead of basing such average attendance upon membership, it were based upon theoretical attendance by every pupil for the full number of days of such school, the actual average attendance would in many cases prove to be not much over one-half the average membership of 24.

The significance of these figures may be best emphasized, perhaps, by a brief summary of them for one year. Thus, in the State of Indiana, for the year 1888, the enumeration of school population was 756,989; the enrollment for the same year, 514,463; average daily attendance, 408,775. The total revenue expended within that year amounted to the sum of $5,235,031. In other words, the average daily attendance was apparently 54 per cent. of the enumeration and 79 per cent. of the enrollment, and the enrollment was

apparently 68 per cent. of the enumeration. But, as a matter of fact, investigation would seem to lead to the conclusion that the average daily attendance, if based upon a theoretical attendance, would show an average of not much more than 25 per cent. of the enumeration and 50 per cent. of the enrollment; and that the enrollment was very considerably exaggerated by the double enrollment of a great many pupils. A careful investigation of the methods of enumeration of children of school age- from 6 to 21-in Indiana, however, has led us to believe that the actual enrollment in the public schools of the State is really much more than 68 per cent. of the real school population of the State. And here enters a factor of injustice to a large number of the school children and tax-payers of the commonwealth that is worthy of careful notice in this connection. In Indiana, as in most of the States of the Union, the law provides that the school revenues of the State shall be apportioned to the several counties and corporations according to the last enumeration therein of children of school age. The manifest purpose of this provision is to make an equitable distribution of such revenues among the communities of the State, so as to secure, as far as possible, a uniform system of common schools, as provided for in the State constitution, including free tuition and terms of equal length throughout the State. Local school authorities may extend the length of their schools by means of local tuition taxes, but it is the purpose of the State that the general apportionment shall be equally made, subject only to such variation as the variety of school needs demand, without injustice to anyone. If the enumeration of children of school age were in each case fairly and conscientiously taken, and its results were not affected by local and improper causes, it might be a reasonably just basis for the apportionment of school revenues; but under the systems followed in most of our States, we question whether it is a proper basis for such distribution. The present tendency of our methods of distribution is to destroy the uniformity of the school system by building up city schools at the expense of the country. While the aggregate attendance in the country schools is better than that of the city schools, and while upon the other hand the city schools possess many economical advantages in matters of organization, yet it is a fact that universally the country-school term is much shorter than that of the cities, while the school taxes of the country are greatly in excess of those in the cities.

By the present system of apportionment of school funds made in the majority of our States, a bid is made for the fraudulent enumerating of children, in order to lighten the burdens of local taxation. This is generally accomplished by the employment of some one to take the enumeration and the payment per capita for the total number enumerated. The enumeration of one of the larger cities in Indiana was recently ordered to be retaken, and a most careful reënumeration resulted in a shortage of nearly five thousand in the aggregate number reported. This makes a difference of more than $16,000 in the amount of State school revenue received annually by such city. Per

haps no more striking evidence of the inequality of school enumeration can be given than the contrasting of cities with the surrounding country. Thus the city of Vincennes, Indiana, has 196 days of school, pays her teachers an average of $4.17 a day, and has accumulated a surplus of $50,000 school tuition fund and yet has never paid any local tuition tax. In the townships of the same county the schools run 83 days less than in the city; the average price paid to teachers is $2.03, and an annual local tuition tax of $5,126 is paid. The city of LaFayette has 190 days of school, pays an average of $3.19 per day to teachers, and pays no local tuition tax. The townships of Tippecanoe county have but an average of 125 days of school, pay an average of $2.27 a day to teachers, and pay $18,950 local tuition tax annually. In some county seats in the State, the State revenue received exceeds $8.50 per capita for the number of pupils enrolled, while in the surrounding townships the average amount is but $3.90 for each pupil enrolled. In the 65 chief cities of Indiana but 19 per cent. of their school revenues is obtained by local tax, while the rest of the State collects 36 per cent. of its school revenues by local taxation. The enumeration of children of school age in these sixtyfive cities, as based upon estimated populations that are, in all probability, judging from past experience, far in excess of the real populations of those cities, is over 36 per cent. of such estimated population; while in the State as a whole, including even these cities, the enumeration, as based upon the population at the time of the decennial taking of the census of the United States, has been barely 30 per cent. Again, the enumeration of the school children in these cities is about 29 per cent. of the aggregate enumeration of the entire State, while their estimated aggregate population is not over 20 per cent. of the population of the State. In Indiana, as in most States, poll tax is levied upon all male citizens between the ages of 21 and 50 years. An exact and perfect record of these polls is made and reported to State authorities. The census statistics show that the number of polls is 24 per cent. of the whole population, while the number of children of school age is 30 per cent. of the population; or, in other words, that the number of school children is one-third greater than the number of polls. Applying this standard to the State of Indiana to-day, it will show that the enumeration of school children in the State is at least 50,000 more than it ought to be. In explanation of the discrepancy, a single example will suffice. In Clark county, exclusive of Jeffersonville, the county seat, the enumeration is 135 per cent. of the poll; while in the city of Jeffersonville the enumeration is 193 per cent. of the poll. In Marion county, within which the capital city of the State is situate, the enumeration, exclusive of Indianapolis, is 138 per cent. of the poll; while in Indianapolis the enumeration is 223 per cent. of the number of polls. Careful investigation leads us to believe that in almost every State having any considerable State school revenue, quite as remarkable discrepancies will be found as we have cited in Indiana.

In the consideration of this point it must not be lost sight of that the child