EXCHANGE 8061 'LZ W 1W AN INVESTIGATION to Determine the Efficiency with which the Compulsory Attendance Law is Enforced IN PHILADELPHIA THESIS PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE UNIVER- SITY OF PENNSYLVANIA IN PARTIAL FULFILLMENT OF THE REQUIRE- MENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BY EDWARD CLINTON BIXLER PHILADELPHIA, PENNSYLVANIA 1913 AN INVESTIGATION to Determine the Efficiency with which the Compulsory Attendance Law is Enforced IN PHILADELPHIA THESIS PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE UNIVER- SITY OF PENNSYLVANIA IN PARTIAL FULFILLMENT OF THE REQUIRE- MENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY BY " EDWARD CLINTON BIXLER PHILADELPHIA. PENNSYLVANIA 1913 ~Ci CHAPTER I. COMPULSORY EDUCATION IN THE UNITED STATES Modern compulsory education has its origin in the desire for universal education at the time that the State was beginning to assume the function of educating the people. The reformers everywhere aimed to perpetuate their faith by educating the people. "The keynote of this attempt was struck by Luther in his address to the councilmen of all the towns of Germany, in 1524. He argued that, if a ruler can compel military service of his subjects, he can also compel the children to attend school as both are for the defence and welfare of the coun- ry." "Calvin, at Geneva, as far as possible, made education obliga- tory. By the close of the sixteenth century the principle of compulsory education had become an essential part of the educational creed of Protestant Germany, and, spreading through other Protestant coun- tries, soon came to be an accomplished fact in some. Holland at- tained a higher standard than could be found anywhere. All, even the lowest classes, could read and write. ''f From Holland and England the movement spread to America. The early New England settlers were educated and in sympathy with the movement for universal education. "The Massachusetts' law of 1642 was a public assertion of the right of the State to educate the child in default of the parent to do his own duty. The Massachu- setts' general court authorized its officials to look into the homes of the people and to unmask parental greed, neglect and abuse; it even as- serted supreme authority to take the child from its parents and edu- cate it at the public expense. "| The Connecticut Colony, in 1650, and the New Haven Colony, in 1655, passed laws containing a provi- sion for obligatory education. These laws were sufficient for their time. "Their successful administration was due primarily to the fav- orable conditions under which they were established. The people were homogeneous and well-to-do. They were intelligent and indus- trious, "t This condition of affairs existed for almost two hundred years until immigration, manufacturing and wealth broke up the form- *Painter, Luther on Education. fPerrin, History of Compulsory Education in New England. JAnnual Report of the United States Commissioner of Education, 1893-94, V. 1, 663. 263986 erly homogeneous people into classes, and the public schools were losing their efficiency, and the system its vitality. "The concentra- tion of the population in the manufacturing and railroad centres aggra- vated the evil. There was less home-work for the children, less op- portunity for parental oversight and control and stronger street temptations. So absenteeism and truancy increased. Thousands of children were brought into the States from England, Ireland and Scot- land, where education was at its lowest ebb children who had never seen inside a schoolroom."* Massachusetts was the first State to direct legislation to the schools to remedy this evil. In 1850 a truant law was passed, follow- ed by a compulsory attendance law in 1852. "This law, like many of the earlier laws of compulsory attendance, attracted little atten- tion, and there is no account of its enforcement, even in the State re- ports."! This law was strengthened by additional acts and amend- ments, especially in 1873 and 1889, so that the towns were not only requested but compelled to make provision for the execution of the law. Since the passage of the Massachusetts' law, other states have directed their attention to prohibiting children from being imposed TABLE I. PROGRESS OF THE ADOPTION OF COMPULSORY SCHOOL ATTENDANCE LAWS. Year Number of Slates and Territories having com- pulsory school attend- ance laws. Population un- der compulsory school attend- ance laws. Population of the United States. Per cent, under compulsory school attend- ance laws. 1870 3 1,919,602 38,558,371 4.98 1875 16 13,412,462 43,700,554 30.69 1880 17 18,414,631 50,155,783 36.72 1885 23 25,898,527 56,221,868 46.06 1890 28 30,920,711 62,622,250 49.38 1895 29 39,970,360 68,748,950 58.14 1900 32 49,394,291 75,994,575 65.00 1905 35 60,156,526 82,584,061 72.84 jThis includes an estimated population of 692,600 subject to com- pulsory school attendance laws in certain counties and cities in Maryland, North Carolina and Tennessee. upon by greedy, ignorant and lazy parents who would put them to work at a very early age. Now thirty-seven states and territories *Martin, Evolution of the Massachusetts Public School System, t Annual Report of the United States Commissioner of Education, 1888-89, V. 1, 471 sq. with the District of Columbia and three states in part, including over seventy per cent of the people of the United States, are under such laws, more or less vigorously enforced. Table I shows the number of states and territories and people where compulsory attendance laws exist, and the population of the United States at the periods named, as given by the United States Commissioner of Education, Re- port for 1906, V. 2, page 1267. With this spread of compulsory school attendance laws, the ques- tion may arise as to what influence they have had upon the enrollment and attendance, both in the United States and in the states having such laws in comparison with those that have not or that have passed them so recently that no satisfactory comparisons can be made. The irregular line in Chart I shows the school enrollment of the United States for every year from 1870-71 to 1906-07. In 1899-00, 20.51 per cent of the people in the United States were enrolled in the schools. A broken line has been drawn to show how many children ought to have been in school in the census years 1880, 1890 and 1900 and 1907 (estimated), if the same per cent of the people had been in school as were enrolled in 1899-00. Since the broken line shows the increase in enrollment based on a corresponding increase in popula- tion, the amount of difference, in numbers, between such an increase and the actual enrollment represented by the irregular line can easily be seen. In Chart II is shown the per cent of the population enrolled in the United States, its five divisions and Pennsylvania, at the periods noted on the chart. The divisions are those used by the Census Bureau and by the United States Commissioner of Education. The chart shows the rapid increase in enrollment in the South Atlantic and South Central Divisions where compulsory attendance laws are not in force, except in a few states. The North Atlantic and North Central Divisions show a steady decline in the per cent of enrollment, and in these divisions every state has had a compulsory attendance law for, at least, ten years. The Western Division gives the best record under compulsory attendance laws, and is the only division which shows no decline since 1900. Possibly the comparison between the South Atlantic and South Central Divisions and the others is not just for the first two,.decades in question because of the reconstruction of their public school systems during the Seventies, and the help given from outside sources. CHART I. NUMBER OF PUPILS ENROLLED IN THE PUPLIC SCHOOLS OF THE UNITED STATES I. Enrollment. II. Estimated Enrollment. CHART II. PER CENT OF THE POPULATION ENROLLED IN THI PUBLIC SCHOOLS United States. II. North Atlantic Division. III. South Atlantic Division. IV. South Central Division. V. North Central Division. VI. Western Division. VII. Pennsylvania. Chart II is misleading in showing any effect of compulsory at- tendance laws that tended to an increase in enrollment, if the effect of the law is to be judged from that standpoint. It does show that in 1907 the states were educating less children, according to population, than in 1900. While this decrease is not satisfactorily explained by the rapid rise in adult immigration, it may be, in part, due to the failure of increased school accommodations to keep pace with the rapid growth of school population. There is, at least, room for doubt that the estimated decrease is due to less efficient administration of compulsory attendance laws. The question may arise whether there are relatively as many of the children in school today as ten or twenty years ago. This can be tested for the census years only. A comparison of the last three censuses will be adequate to the purpose. According to the census of 1880, 34.3 per cent of the population was 5 to 19 years inclusive. In 1879-80, 19.67 per cent of the population and 57.3 per cent of the children 5 to 19 years inclusive were enrolled in the public schools. In 1890 the per cent of children 5 to 19 years inclusive had decreased to 33.8 per cent of the population. The enrollment for 1889-90 had increased to 20.32 per cent of the population and to 60.1 per cent of the children 5 to 19 years inclusive. In 1900 the per cent of child- ren 5 to 19 years inclusive had decreased to 32.3 per cent of the pop- ulation. The enrollment for 1899-00 had increased to 20.51 per cent of the population and to 63.5 per cent of the children 5 to 19 years in- clusive. These figures, which are the only ones upon which any just and accurate conclusion can be based, show an increase in enrollment, in proportion to the population of school age, for the whole period, which is steady for the census years. It is not possible to say whether a decrease in the number of children 5 to 19 years inclusive, in propor- tion to the population, explains the steady decrease in the per cent of population enrolled during the years since the last census, although this seems to be a plausible explanation. But since no rate of decrease in the per cent of children 5 to 19 years inclusive can be established with any degree of accuracy, and any estimation of the proportion of such children enrolled could therefore not be expressed in figures, with safety, all conclusions as to the probable number or per cent are indeterminable. Chart III shows the per cent of the population 5 to 1 8 years en- rolled in school, for the United States, its divisions and Pennsylvania, at the periods noted on the chart, from figures based on the Report of the United States Commissioner of Education, 1907, V. II, page 549. In considering the charts of this chapter it must be held in mind that almost all the states of the South Atlantic and South Central Divisions have no compulsory attendance laws. For the period 1899-00 to 1906-07 this chart does not present a very favorable record for any division except the Western. No explanation for this can be given, unless, as stated above, the next census may reveal a much lower num- ber of children 5 to 19 years inclusive or 5 to 1 8 years, in proportion to population, or show that the private schools are increasing in num- bers. The school censuses of the states ought to give some basis for proving the possible correctness of the figures given above. Some di- visions show a lower per cent of the number of children 5 to 18 years enrolled in 1906-07 than they did in 1870-71. This is true of Penn- sylvania which shows a steady decline since 1871. Since it is both of interest and importance for the problem under consideration to find out what opportunities are offered to the children to obtain an education as well as to show what has actually been ac- complished in numbers, the length of the school year has been indi- cated on a chart, from figures based on the Report of the United States Commissioner of Education, 1907, V. II. page 553. If the school year is lengthened, there may be a tendency to drop out earlier. The children may become tired of school or the parents may decide that the children have received sufficient education for the work they may do in life, at an earlier age, which they accomplished, in part, through a longer school year. In Chart IV the length of the school year, in days, is given for the United States, its five divisions and Pennsylvania. It is gratifying to discern that there has been a steady increase in the length of the school year since 1880. In view of this fact it will be interesting to note how much more schooling is given to each child enrolled, each year, in terms of days. If our educational system is undergoing improvements and greater opportunities are offered for gaining an education, it is certainly im- portant to know what has been the result of these advantages. Since this is not obtainable in statistics which indicate the amount of knowledge and training gained by the child, it is essential to, at least, make a comparison in days of actual attendance. This is shown in Chart V for the years given at the top of the chart. This chart gives a favorable report for the United States and all its divisions, and es- CHART III. PER CENT OF THE POPULATION 5 TO 18 YEARS OF AGE ENROLLED IN THE PUBLIC SCHOOLS. CHART IV. LENGTH OF THE SCHOOL YEAR. I. United States. V. II. North Atlantic Division. VI. III. South Atlantic Division. VII IV. South Central Division. North Central Division. Western Division. Pennsylvania. pecially for those divisions which showed a decrease in the preceding charts, even under the influence of the compulsory attendance laws. Even the increase, in the divisions not having compulsory attendance laws, in the enrollment shown in Charts II and III does not make such a good showing when measured in terms of days schooling a year for each child enrolled. In Chart VI this matter has been presented in a slightly dif- ferent way. Here the entire population of the United States, its different divisions and Pennsylvania have been used as a basis of com- putation to test what has actually been done by our educational sys- tem, in terms of days. In this chart great variations are seen. Almost all the divisions are increasing the amount of schooling given in pro- portion to population. This is most marked in the Western Division, but, excepting this division, the change in amount has not been so marked since 1900. Only one division under compulsory attendance laws shows a decrease. The above charts give some idea as to what our educational sys- tem has accomplished during the years 1870-71 to 1906-07. It is dis- appointing in that the child does not receive more days schooling each year, considering the increase in the length of the school year, and also in the fact that there is not a larger per cent of the children enrolled in school. The number of illiterates, ten years of age and over, is also large for the United States and especially for some divisions. In Chart VII the per cent of the population, ten years of age and over, illiterate, for the last three censuses, is shown. The decrease in per cent is marked for almost all the divisions, and Pennsylvania shows a slight decrease. A comparison between the different divisions, in amount of illiteracy, is unfair, because in the two Southern divisions the problem of educat- ing the negro still exists, and, in other divisions, especially where there are large manufacturing and mining interests, the per cent of foreign illiterate population is very large. This is especially true of Penn- sylvania where 64 per cent of the illiterates are foreign born, while the number for the 'United States is 22 per cent. Negro illiterates, however, made up only 7 per cent of the illiterates of Pennsylvania and 46 per cent for the United States. From this it can readily be seen that a state which has few negroes and foreigners should have a small number of illiterates. In Table II the per cent of illiteracy for the six largest cities of 10 CHART V. AVERAGE NUMBER OF DAYS ATTENDED EACH YEAR BY EACH PUPIL ENROLLED IN THE PUBLIC SCHOOLS. CHART VI. DAYS IN SCHOOL EACH YEAR, BASED ON POPULATION I. United States. II. North Atlantic Division. III. South Atlantic Division. IV. South Central Division. V. North Central Division. VI. Western Division. VII. Pennsylvania. VIII. Philadelphia. 11 CHART VII. PER CENT OF THE POPULATION, TEN YEARS OF AGE AND OVER, ILLITERATE. TABLE II. PER CENT OF THE POPULATION, TEN YEARS OF AGE AND OVER, ILLITERATE. City All Native White Foreign White \ Negro 1900 1890 1900 1900 1900 New York Philadelphia Chicago St. Louis Boston Baltimore 6.8 4.4 3.9 4.4 5.1 7.2 7.69 4.97 4.63 5.89 5.69 9.80 0.4 0.6 0.2 0.9 0.2 1.3 13.9 12.1 8.2 9.8 11.3 12.9 10.0 11.8 8.0 21.3 10.6 25.8 the United States has been given for the census years 1890 and 1900. The position of Philadelphia among the cities can readily be seen from the table. The columns headed foreign white and negro show clearly the effect immigration and the negro race have in increasing the per cent of illiteracy. But the problem that is confronting the school boards and public is the large army of children of school age that are not enrolled in school, many of whom, if neglected, will help to swell the number of illiterates for future decades. In Table III these facts have been presented for the census year 1900, as given by the United States cen- sus. This table shows that 20.16 per cent of the children 10 to 14 years inclusive were not enrolled in school in 1900. 577,649, or over 9 per cent, of the illiterates in the United States were between the years 10 to 14 inclusive in 1900. The per cent of non-attendance 12 TABLE III. NUMBER OF CHILDREN IN SCHOOL AND PER CENT NOT ENROLLED IN SCHOOL. State, Territory or City. 5 to 9 years inclusive 10 to 14 years inclusive Census Enrolled in school c^l i 19.7 18.7% 10 11 12 13 14 21.5 19.6 20.3 19.2 19.4 21.4 18.1 21.4 18.9 20.2 22.4 18.5 21.6 18.8 18.7 21.7% 18.7H 21.1 18.9% 19.4^ 21.1 19.9 19.8 19.5 19.7 21.0 18.7 20.8 19.1 20.4 21.7 18.8 21.1 19.3 19.1 21.2% 19.1^ 20.5% 19.3 19.7% 15 16 17 18 19 20.3 20.6 19.7 20.3 19.1 19.7 21.2 19.1 21.4 18.6 18.7 19.7 18.9 22.6 20.1 19.5% 20.5 19.2% 21.4% 19.2% 20.0 20.5 19.7 20.2 19.6 19.2 21.1 19.7 20.8 19.2 19.1 19.7 19.6 21.4 20.2 19.4% 20.4^, 19.6% 20.8 19.6% 20 20.8 20.7 21.9 21.1% 20.1 20.2 21.4 20.5% possible, Table VI has been formed. This table shows what per cent of the children in a five year period belong to each year. The per cents have been worked out both for the United States and for Penn- sylvania and have been computed for the last three census years. This plan seemed preferable to taking one-fifth of those in any five year period as representative of the number that belonged to a certain year of that period. It will be seen from the table that the amount of difference between the years of each period is very small, both for the United States and for Pennsylvania. Yet, for the sake of accuracy, this should be considered. For Philadelphia and the cities of Penn- sylvania the per cents of Pennsylvania given under 1900 will be used in determining the number of children of the ages to be used in the comparisons. In all other comparisons the same year for the United States will furnish the per cents. Since there is no school census reported for Philadelphia for 1900, a mean between the 1899 and 1901 school censuses has been taken to represent the census for 1900. Comparisons have been made between this mean and the national census for 1900, for the age pe- riods given in Table VII. This table shows that the difference be- tween the national census and the school census (mean between the 1899 and 1901 school censuses taken by the attendance officers) in- creases as it reaches the age period where childrn leave school to go 31 TABLE VII. COMPARISONS OF SCHOOL AND NATIONAL CENSUSES OF PHILADELPHIA FOR IQOO. Age Period School census (mean between 1899 and 1901 school censuses) Number bas- ed on nation- al census. Amount the school census is below the national census. Amount of variat i o n by years. Number Per Cent. 6 to 8 years 6 to 16 " 8 to 16 " 16 to 21 " 6 to 21 " 47,373 205,081 157,708 68,334 273,415 51,041 231,791 180,750 114,264 346,055 3,668 26,710 23,042 45,930 72,640 7.19 11.52 12.75 40.20 20.99 1,834 2,671 2,880 9,186 4,843 to work. The difference is especially large in the 1 6 to 21 years' pe- riod. In computing the per cent of difference in column 5, the num- bers in column 3 have been used as the bases. Table VIII has been constructed to show what per cent of the population of Philadelphia was of the different age periods given in the table, for 1900. The per cents for Philadelphia have been determined from the mean of the 1*899 and 1901 school censuses, taken by the attendance officers, and from the national cen- sus. The total population, upon which these per cents have been TABLE VIII. PER CENT OF THE POPULATION OF DIFFERENT AGE PERIODS IN I9OO. Age Peroid Philadelphia Pennsylvania United States School census National census 6 to 8 years 6 to 16 " 8 to 16 " 16 to 21 " 6 to 21 " 3.66 15.85 12.19 5.28 21.13 3.95 17.92 13.97 8.83 26.75 4.44 20.49 16.05 9.53 30.02 4.78 22.06 17.28 10.02 32.08 based, is from the national census for 1900. To these statistics have been added the per cents for Pennsylvania and the United States, for the same age periods. In this table the per cents of the school census are below the per cents determined from the national census, for Philadelphia. The amount of difference increases to the age for go- ing to work. The table shows that a comparison between a city and state, or the United States, can not be made with any due degree of accuracy. In order to compare Philadelphia with other cities, Table IX has been computed. Instead of the age periods used in Table VIII, five year periods have ben used. Nine cities have been chosen. The per cent of children of the different five year periods, 32 TABLE IX. COMPARISONS OF THE PER CENT OF THE POPULATION OF DIFFERENT AGE PERIODS IN IQOO. CITY OR STATE 5 to 9 years 10 to 14 years 15 to 19 years 20 to 24 years 5 to 19 years New York 10.32 8.76 8.81 10.32 27.89 Boston Buffalo 8.72 11.34 7.55 10.17 7.53 9.15 10.39 9.28 23.80 30.66 Chester 10.33 9.26 9.37 10.72 28.96 Cincinnati 9.87 9.39 9.59 10.38 28.85 Cleveland 10.96 9.47 9.54 10.38 29.97 Fort \Vayne 10.28 9.84 9.99 10.01 30.11 Harrisburg 9.49 9.33 9.49 10.49 28.31 Philadelphia . . . 9.69 8.49 8.49 10.17 26.67 Average 10.11 9.14 9.11 10.24 28.36 Massachusetts '..... 9.13 8.17 8.48 9.96 25.78 New York 9.82 8.84 8.81 9.67 27.47 Ohio 10.48 9.98 9.81 9.46 30.27 Pennsylvania 10.89 9.90 9.51 9.59 30. bO Average 10.08 9.22 9.15 9.67 28.45 United States 11.70 10.70 9.90 9.70 32.30 from 5 to 24 years inclusive, have been determined from the national census for these cities. To these statistics have been added those for four states and the United States. As the 5 to 19 years' inclusive period covers the years for which the school censuses are generally taken, the variations between the cities for this period will be noted. Boston has the lowest and Buffalo, the highest per cent of the population 5 to 19 years inclusive. The amount of difference between these two cities is not quite 7 per cent. Philadelphia has second place in the 5 to 19 years' inclusive column, in smallness of per cent. It is also almost two per cent be- low the average for the nine cities. Notwithstanding this fact, the per cent is much too high when a comparison is made with the school census taken by the attendance officers. Table VIII showed that 21.13 per cent of the population of Philadelphia in 1900 was be- tween 6 and 21 years, according to the mean between the school cen- suses for 1899 and 1901. Table IX shows that 26.67 per cent of the population was between 5 and 19 years inclusive in 1900, according to the national census. If the year 20 in the 6 to 21 years' period were exchanged for the year 5 in the 5 to 19 years' inclusive period, the former per cent would be only a fraction higher, 26.75 per cent. The 21.13 per cent, however, is lower than any per cent given in the 5 to 19 years' inclusive period. Even Boston, where the difference between the school and national censuses was very small for 1900, 33 as will be shown below, has a higher per cent than the 21.13 per cent of the school census of Philadelphia. The proportion of the population 5 to 19 years inclusive is, as a rule, higher for the states than the cities given in the above table. There is variation between the states as is the case with the cities. The United States shows a higher per cent for the 5 to 19 years' inclusive period than any of the cities or states of Table IX. In Fort Wayne, Indiana, the Superintendent of schools has made comparisons between the school census of that City and the national census.* He has accepted 25 per cent as the minimum and 30 per cent as the maximum limits of the population between 6 and 21 years. If the school census shows a smaller number than 25 per cent of the population of the city, according to the national census, or a number larger than 30 per cent, the accuracy of the school census is open to question. Possibly these limits do not permit of enough variation to be applied to any city. Maximum and minimum limits would have to be determined for each state and city, as can readily be seen by a reference to Table IX. Immigration and the size of the city would have to be considered. To include all the cities of Table IX, the minimum and maximum limits would have to be 23 and 33 per cent respectively. For Pennsylvania the minimum and maximum limits of the per cent of the population between 5 and 19 years inclusive or 6 and 21 years might be placed at 25 and 31 per cent respectively. For Philadelphia the limits suggested for Fort Wayne, 25 and 30 per cent, appear to be satisfactory. This mini- mum limit of 25 per cent, however, is too high for the school census of 21.13 per cent given in Table VIII. The above comparisons between school and national censuses have been made on the assumption that the national census is correct. The charge of incorrectness may be made against the national census, since there are difficulties to overcome in taking the national as well as the school censuses. But, since the national census makes an enum- eration of all the people and the school census records only those 6 to 1 6 years in Pennsylvania ( formerly 6 to 21 years), accuracy ought to be in favor of the former. There would be less reason for eluding or giving a false or inaccurate statement to the national census officers, *Annual Report of the Superintendent of Public Schools of Fort Wayne, Indiana, 1907, pages 17-18. 34 since attendance or non-attendance at school would not be affected by any report or information given. In order to test the accuracy of the national census, a comparison has been made between the number of children of a certain age report- ed in school in Philadelphia, for 1900, by the national census, and the number reported in school for the same age and period by the school reports of the City. The national census for 1900 reported 165,818 children 5 to 14 years inclusive enrolled in the schools of Philadelphia. This number included those enrolled in the private as well as in the public schools. It has not been possible to determine what per cent of the children were enrolled in the public schools. According to the Report of the Bureau of Compulsory Education for 1905, page 9> 75-27 per cent of the children 6 to 16 years enrolled were in the public schools. According to the Report of the same Bureau for 1908, page 14, 73.74 per cent of the children 6 to 16 years enrolled were in the public schools. Assuming the accuracy of these figures and that the variation for the period 1900 to 1905 was very small, 75 per cent can be accepted as a generous basis for the per cent of child- ren 6 to 1 6 years enrolled in the public schools in 1900, out of the total number 6 to 16 years enrolled in all the schools. This is slightly lower than the per cent for 1905, although a higher per cent would favor the comparison. The next difficulty encountered in this comparison was to determine how many children enrolled in the public schools were between 5 and 14 years inclusive. This informa- tion could be obtained for June 3Oth, 1908,* when 87.62 per cent of the children in the public schools were between 5 and 14 years inclusive. If 75 per cent of the number reported by the national census were in the public schools, there were 124,364 children 5 to 14 years inclusive in the public schools when the census was taken. Assuming that 87.62 per cent of the children in the public schools on June 3Oth, 1900, were 5 to 14 years inclusive, as was the case on June 3Oth, 1908, and that 124,364 represented this per cent, the whole number in the public schools on June 3Oth, 1900, should have been 141,935. The enrollment for June 3Oth is the lowest of any month of the year. It was e'qual to 94.95 per cent of the average enrollment for the year 1908. Assuming that the difference between the June 3Oth and average enrollment has been very small, the 94.95 *Annual Report of the Superintendent of Public Schools of Philadelphia, 1908 35 per cent has been accepted for 1900. If 141,935 represented the enrollment on June 3Oth, 1900, and this was 94.95 per cent of the average enrollment, the average number of children enrolled in 1900 was 149,468. The School Report for Philadelphia for 1900 gave an average enrollment of 147,314. The amount of difference be- tween the average enrollment determined from the national census by the aid of the per cents based on the school reports and the average enrollment reported by the Superintendent of Schools for 1900 is 2,154, or 1.4 per cent based on the estimated average enrollment. The per cent of difference between the national census and the school reports, when enrollments are compared, is very much smaller than the difference between the national and school censuses. In the case of the enrollment comparisons, the difference may be ac- counted for by a difference in the time of reporting the enrollments compared, since one month would give a difference, in the case of some months, of more than the per cent of difference given above, that is, the 1.4 per cent. If the same per cents were used to compute the 1908 average enrollment from the national census, the resulting number would be below the average enrollment given in the School Report of Philadelphia for 1908. Since the national census would have had to be estimated for that year, the comparison has not been recorded in this investigation. The above comparisons favor the as- sumption that the national census is correct. The enrollment must be accepted as a safe basis by which to test the accuracy of the national census. The national census, therefore, can be accepted as correct in the comparisons already made between the national and school cen- suses of Philadelphia and similar comparisons that are now to be made for other cities. In Table X comparisons have been made between the school and the national censuses for 1900 for thirteen cities of the United States. The fourth column shows how many children were of a certain age, for the respective cities, determined from the na- tional census for 1900. Since the national census gives five year periods only, in the case of cities, to show what part of the whole population is of a certain age, the per cents in Table VI have been used in computing part periods. Chicago may be taken to illustrate the method employed. The school census of Chicago reported the num- ber of children 6 to 21 years. Therefore, the number for the two periods of the national census, 9 to 19 years inclusive, can be accept- 36 TABLE X. COMPARISONS OF SCHOOL AND NATIONAL CENSUSES FOR I9OO. City Age Period School census Number based on national census School census above national census. National cen- sus above school census. Number percent number percent Boston . . . 5-15 years 90,144 91,258 1,114 1.22 Chicago . . 6-21 626,516 490,913 135,603 27.62 Cincinnati . 6-21 108,391 94,490 13,901 14.71 Cleveland 6-21 109,047 114,122 5,075 4.45 Detroit . . . 5-20 81,681 86,188 4,507 5.23 Fort Wayne . 6-21 14,036 13,576 460 3.39 Indianapolis . 6-21 40,073 45,778 5,705 12.46 Philadelphia 6-21 273,415 346,055 72,640 20.99 Allentown . 6-21 8,974 10,282 1,488 14.47 Chester . . . 6-21 6,886 10,858 3,972 36.58 Erie .... 6-21 16,195 15,439 756 4.90 Harrisburg . 6-21 11,296 14,288 2,992 20.94 Lancaster . . 6-21 9,999 12,175 2,176 17.87 ted without modification. To this number was added the number of children 6 to 9 years by taking 20.4 per cent from the 5 to 9 years' inclusive period, according to Table VI, column 2, under the United States. Ln the same way the number 20 years old can be deter- mined from the national census. The sum of those 6 to 9 years in- clusive, 10 to 14 years inclusive, 15 to 19 years inclusive, and 2O years, was the number of children in Chicago 6 to 21 years, in 1900, according to the national census. The school censuses given in col- umn 3 have been taken from the Report of the United States Com- missioner of Education for 1900. The amount and per cent of difference between the school and national censuses have been shown in the table. In the cities not in Pennsylvania, this is greatest in Chicago, where the school census was 27.62 per cent higher than the number based on the national cen- sus. According to the school census of Chicago, 36.88 per cent of the population was 6 to 20 years inclusive. This per cent is above any of the per cents in the 5 to 19 years' inclusive column of Table IX. Especial attention should be called to Boston and Detroit where the school and national censuses could be compared without any esti- mates. The per cent of difference between the two censuses is very small, especially in the case of Boston. The difference in the time of taking the two censuses could account for the difference in Boston. Philadelphia shows a school census lower than the number based on 37 the national census. This appears to be another point in favor of the argument that the school census of Philadelphia, taken by the attend- ance officers, has been too low, and a table below will show that it is still too low. A comparison between the six cities in Pennsylvania shows that the conditions are not very promising in the State. The 36.58 per cent for Chester is considered correct in the light of a special census that added the names of almost 2,000 children to the number enrolled by the ward assessors. TABLE XI. COMPARISONS OF ESTIMATED AND NATIONAL CENSUSES. CITY OR STATE 1900 1890 National census Estimated census Estimates too low National census Estimated census Estimates too low Amount Per Cent Amount Per Cent Boston Buffalo Cincinnati Cleveland Washington Philadelphia Chicago 560,892 352,387 325,902 381,768 278,718 1,293,697 1,698,575 534,115 356,194 338,677 362,560 283,160 1,246,758 1,696,515 26,777 3,807* 12,775* 19,208 4,442* 46,939 2,060 4.8 1.1* 3.9* 5.0 1.6* 3.6 0.1 448,477 255,664 296,908 261,353 230,392 1,046,964 475,152 192,554 294,039 227,463 246,049 1,020,318 26,675* 63,100 2,869 33,890 15,657* 26,646 6.0* 24.7 1.0 13.0 6.8* 2.5 Massachusetts New York Ohio Pennsylvania 2,805,346 7,268,894 4,157,545 6,302,115 2,694,801 6,912,835 4,146,570 6,233,137 110,545 356,059 10,975 68,978 3.9 4.9 0.3 1.1 2,238,943 5,997,853 3,672,316 5,258,014 2,108,819 5,782,983 3,730,864 5,043,831 130,124 214,870 58,548* 214,183 5.8 3.6 1.6* 4.1 1.4 United States 75,693,734 75,088,717 605,017 0.8 62,622,250 61,753,195 869,055 *Estimates too high In order to compare the school censuses of Philadel- phia, for the years 1901 to 1908, with the national census, it will be necessary to accept or make estimates for these years, based on the national census for 1900. In order to prove the correctness of these estimates, the censuses of seven cities, four states and the United States have been estimated for 1890 and 1900. To estimate the population of a city or state for 1900, the difference between the cen- suses of said city or state for 1890 and 1880 was added to the 1890 census. This number has been recorded in Table XI in the column headed "estimated census." In another column of the table the national census for 1900 has been given for the same city or state. In the same way estimates were made for 1890. The difference be- tween the national and estimatd censuses has been shown in the table. The cities have been chosen largely on the basis of their size. Older cities have been chosen because they generally show less fluctation in 38 the per cent of increase in population. This was considered legiti- mate from the fact that Philadelphia is in this class. Table XI shows clearly that the difference between the na- tional and estimated censuses is small in the case of most of the cities and states. The estimate is more often too low than too high. In Philadelphia and in Pennsylvania the estimates are lower than the actual censuses for both years compared. Since the per cent of differ- ence is not great for either and nearly the same for the two years compared, the estimates below for the years 1901 to 1908 must be considered almost correct. If there is any variation from the actual numbers, which the census for 1910 will show, the comparisons just made would favor the assumption that the estimates are too low. If the estimates are too low, the comparisons between the school and estimated censuses will be more favorable to the former in Table XII. TABLE XII. :OMPARISONS OF THE SCHOOL AND ESTIMATED NATIONAL CENSUSES OF PHILADELPHIA. Year School census Estimated from national census Per cent for estimate Difference Amount Per cent 1901 208,570 235,988 17.90 27,418 11.6 1902 204,423 240,270 17.89 35,847 14.9 1903 212,308 244,547 17.88 32,239 13.2 1904 217,935 248,820 17.87 30,885 12.4 1905 223,591 253,087 17.86 29,496 11.7 1906 227,370 257,350 17.85 29,980 11.6 1907 233,084 261,608 17.84 28,524 10.9 1908 236,838 265,860 17.83 29,022 10.9 In Table XII the school census has been compared with the estimated national census for Philadelphia, for the age period 6 to 16 years, for the years 1901 to 1908. In estimating the relative number of children 6 to 16 years, due allowance has been made for a decrease in the number of children, which was found to be o.i per cent for the period from 1890 to 1900. This makes a de- crease of o.oi per cent a year. Column 4 of Table XII shows this decrease in the per cents. This per cent of decrease is very small and changes the estimates very little. It has been taken into ac- count, however, in order to make the comparisons as exact as possible and just for the school census. The school census for 1902 is less than that for 1901, which is explained in the reports as due to a difference in the time of the year when the census was taken. This made it more difficult to en- 39 roll the full number in 1902. Possibly similar difficulties are encoun- tered at other times, which may, in part, account for the difference between the school and estimated censuses of Table XII. The in- crease in the school census since 1903 has been almost in proportion to the estimated increase in the population of the City. But, if the estimated census is found to be too low, when a comparison can be made with the next national census, the school census will be proved to be less accurate for each year that it is taken, unless the yearly gain is greatly increased. In Table XII the per cents of difference are based on the estimated and not the school census. These per cents do not vary much during the years compared, excluding 1902. TABLE XIII. COMPARISONS OF SCHOOL AND ESTIMATED NATIONAL CENSUSES. City Year Age Period School census Estimated from national census Amount school census is below national cen- sus. Number Per cent Buffalo .... Boston .... Cleveland . . Harrisburg . . 1906 1907 1907 1907 5-18 years 5-15 " 6-21 " 6-21 " 84,530 104,150 125,368 8,595 108,556 104,443 135,941 10,450 24,026 293 10,573 1,855 22.1 0.3 7.8 17.8 In Table XIII comparisons have been made between the school and estimated censuses for four cities, for the years given in the second column of the table. The school census was obtained from the published reports of the respective cities, for the years compared. The comparison for Boston is of interest. The difference between the school and estimated census is only 0.3 per cent. However, if the esti- mated census is based on the state census of 1905, the difference between the school and estimated census is 4.9 per cent. In the case of Buffalo the estimate based on the state census for 1905 is lower than the estimate based on the national census. Table XIII only emphasizes the fact that there is difficulty jn obtaining an accurate school census. The problem concerns many ci- ties. The comparisons in this chapter between school and national censuses, for the same age periods, have been made to show the approximate amount of difference between the two censuses. Every means has been employed to make fair and accurate comparisons. All influences that might modify the results, however small, have been taken into account. The above comparisons seem to give rather con- clusive evidence that the school census for Philadelphia is too low. 40 The great variation in Table VII can be explained in no other way than from the fact that, as the children approach the age to go to work, there is a tendency to give false reports to the attendance officers when taking the census, escape enrollment in the census or the proper effort is not made to make a complete census. While Table XII may not give the exact amount that the school census lacks of the whole number, it can be taken as approximately correct. It does show, however, the need of a better organized and more efficient system for taking the school census. CHAPTER V. ATTENDANCE RECORDS. INVESTIGATION OF THE ENFORCEMENT OF THE COMPULSORY ATTENDANCE LAW IN THE SCHOOLS OF PHILADELPHIA. An investigation was made of the cards used to report absentees to the attendance officers, to ascertain the number of times each one was reported, the period of each report and the character of the rea- sons recorded by the attendance officers for the absences. All the attendance cards of several schools were examined. These schools were considered representative of the conditions existing in Philadel- phia, so that the results of this investigation show how accurately these records are kept in the schools and how the compulsory attend- ance law is respected by the teachers in reporting absentees. The work was frought with difficulties because of the negligent manner often employed in recording the data found on the cards. Doubtless the data were intelligible to those who made the entries, yet the cards were not open to easy and definite investigation. By a process of com- parisons of the various facts recorded on the cards, the statistics ob- tained and shown in the following tables, are accurate. The special difficulties of this investigation will be noted under the tables below, which give the statistics for which the data were not easily obtainable. This investigation was made in February, 1909, immediately after the semi-annual promotions. The statistics cover a period of one year, from February 1st, 1908 to February ist, 1909. In Table XIV is given the total enrollment of all the schools investigated. For the purpose of further comparisons, the statistics for the boys and the girls are given separately. 41 TABLE XIV. TOTAL ENROLLMENT OF SCHOOLS INVESTIGATED, BY SEX, GRADE AND AGE. BOYS GIRLS AGE 1 TJ ^ x x P 3 1 | "u - f 3 11 1 J3 P I = i 1 - & j. - fa 11 1 i 5 3 3 9 h 2 11 6 50 2 52 47 23 70 7 36 31 3 71 39 35 2 76 8 19 34 26 1 1 81 24 36 41 2 103 9 1 22 53 33 1 110 2 15 67 26 2 11?, 10 6 31 45 32 7 121 2 6 42 38 32 4 11 5 28 34 38 28 3 136 4 26 54 59 30 3 i 177 12 1 14 37 42 46 24 3 167 13 34 57 44 29 i 178 13 3 6 17 30 45 31 14 146 7 24 31 33 39 15 149 14 1 9 19 12 27 28 96 3 7 13 28 23 22 96 15 2 1 2 4 9 12 30 1 3 12 13 11 40 16 2 2 8 12 1 1 2 6 6 16 17 1 1 2 4 1 4 5 Total 109 104 164 177 165 145 97 67 1028 123 121 201 188 198 153 117 56 1157 TABLE XV. NUMBER REPORTED TO THE ATTENDANCE OFFICERS IN SCHOOLS INVESTIGATED, BY SEX, GRADE AND AGE. BOYS GIRLS AGE fa Second 1 A c 1 x 1 X id I i Second J I X fa x c: 11 | \ i 1 8 1 2 2 5 3 5 3 11 9 7 9 16 5 6 1 12 10 1 5 3 2 1 12 1 5 6 5 3 20 11 3 11 3 10 1 28 7 10 9 4 30 12 2 14 9 18 1 44 3 12 19 11 4 49 13 .1 2 10 9 17 7 46 4 6 19 19 5 i 54 14 1 7 5 5 1 19 2 4 7 4 17 15 1 1 1 3 1 1 8 2 5 2 9 16 1 1 1 i 4 Total 1 H~ -32 32 38 46 14 2 179 4 15 29 36 57 48 15 2 206 42 In Table XV is shown the sex, grade and age of the pupils re- ported to the attendance officers from February* ist, 1908 to Febru- ary ist, 1909, for the schools investigated. The table shows that the number reported increased from the first to the sixth grade in the case of the boys and to the fifth grade in the case of the girls. Then there was a large yearly decrease to the end of the eigth grade. Of course, some were reported in the grade below the one in which they were enrolled when the investigation was conducted. This was due to the fact that the records were made from the cards after the semi-annual promotions. It was not possible to trace each pupil's record to the grade where the absence or absences occurred. The column headed totals shows how many boys and girls reported were of a certain age. The number reported increased from the eighth to the thirteenth year. Then there was a large annual decrease to the end of the compulsory school age. The boys and girls who have been recorded in the table as sixteen years of age, were reported be- fore they had reached the compulsory age limit. Their record was in- cluded because the absences for which they were reported occurred during the year for which the investigation was made and before they had reached their sixteenth birthday. The difference between the number of boys and girls, reported at each age period, is very small. The per cent of the whole number of boys and girls reported shows almost no difference. 17.4 per cent of the whole number of the boys and 17.8 per cent of the whole number of the girls, in the schools investigated, were reported to the attendance officers. It ,was not possible, however, to determine exactly what per cent of the pupils were reported in one year. This was due to the fact that the attend- ance cards of those who have quit school or moved to another district or city were withdrawn from the teachers' records or were supposed to be withdrawn. In a number of cases, however, these cards were still in the possession of the teachers under whom these children had been in school. In order to avoid any difficulty from including the record of children not in school, the attendance cards were compared with the enrollment cards. In some cases the grade of the children was thus determined. This comparison made the age records more nearly accurate. Sometimes the age record on the attendance and enrollment cards did not agree. The latter was always accepted as correct, especially when it showed a corrected age which differed from 43 the first recorded age in the year and occasionally in the month and day. Table XVI shows the number of days for which the pupils were reported each time, by sex and grade. This table shows how many re- ports in each grade were for each of the number of days given in col- umn i. The total number of reports for any period of days is given in the last column under totals. Since the compulsory at- tendance law requires that each child who has been absent three days or their equivalent, without a lawful excuse, be reported to the attendance officers, especial attention was paid to the two and one-half and three and one-half days. When the period of absence was any other day plus a half day, the half day was counted as a whole day in order not to burden the table with extra details. This has not made the table less valuable, since the aim has been to show that the period for which absences are reported is not always three days or their equivalent, as the law of compulsory attendance requires. Difficulty was encounted in determining the number of days for which children were reported each time, due to the way in which the data were recorded. Sometimes only the two days which gave the limits of the period of absence were recorded. Some cards showed a separation of the days reported by a dash, and when this was used for two days, it was necessary to study the teacher's method of recording dates before deciding the length of the period reported. All ab- sences extending over nine days were classed under miscellaneous. These generally represented periods of absence due to sickness. School days only were counted in determining the length of the absences. This table shows that there is no uniform rule followed in re- porting children who are absent. No reason has been found to satis- factorily explain the great variation in number of days for which absences are reported. It certainly shows a disregard for the com- pulsory attendance law or ignorance of its demands. The manner in which the record of days absent is kept on the attendance cards lends weight to the argument that teachers are negligent in performing their part in the enforcement of the compulsory attendance law. There is a possible explanation, however, of the periods under three days. In many cases this may represent a second report for the same period of absence. If the day recorded on the card, which represent- ed the first day absent for which the report was made, was the next 44 TABLE XVI. NUMBER OF DAYS FOR WHICH PUPILS WERE REPORTED EACH TIME, BY SEX AND GRADE. BOYS GIRLS 43 *o DAYS I 1 i Fourth 35 Sevent 43 bJ> 2 i 1 a o I 43 P 43 i c ^ Is 1 i I | 1 i 11 8 10 1 32 1 1 i 1 12 16 2 19 13 14 11 10 3 70 16 10 17 17 8 5 i 74 5 2 3 1 11 1 3 3 1 8 3 2 23 16 18 18 18 5 98 10 13 15 23 20 5 3 89 1 1 2 4 1 3 4 4 2 17 9 18 22 6 3 75 1 9 16 14 17 9 4 1 71 5 13 18 12 8 13 3 1 68 2 10 22 19 22 13 5 93 6 2 5 3 7 13 3 1 34 2 3 4 17 5 3 1 35 7 2 5 4 5 4 2 22 1 5 3 6 7 7 2 1 32 8 1 2 3 4 1 2 13 3 1 3 5 2 14 9 1 1 3 3 8 2 3 4 3 12 Miscel- 1 3 4 13 10 17 2 50 1 6 11 12 22 14 3 69 laneous TABLE XVII. NUMBER OF TIMES EACH CHILD WAS REPORTED, BY SEX AND GRADE. BOYS GIRLS TIMES I to Second 3 lif P ^ 1 "x 33 Seventh 43 I W 1 | Second d IH 43 P Fourth 43 rS c/5 Seventh 43 5 a 1 1 1 3 17 11 19 26 10 2 89 2 7 11 12 26 27 10 95 2 1 6 8 7 9 31 1 2 10 10 12 9 1 i 46 3 1 3 2 5 3 2 16 1 1 4 5 5 3 2 21 4 2 1 4 2 3 12 1 4 5 4 14 5 2 1 1 2 1 7 2 4 4 1 i 12 6 2 1 2 1 6 2 1 3 1 1 8 7 1 1 2 1 1 6 1 1 8 2 1 3 1 2 3 9 1 1 2 1 1 10-15 3 1 2 1 7 2 3 5 45 school day following the last day reported in the previous report for absence, the conclusion can be made that the first investigation had failed to yield the desired result. If the first or only day recorded in this second report is two or more school days after the last day report- ed in the previous record of days absent, there are two possible solu- tions for a one or two days' report. The children may not have re- turned after the previous investigation. In making a second report oi the same period, the teacher may have considered it sufficient to record only the day on which the second report was made because of the fact that the limit of three days had been reached. The attend- ance officer was to assume from this that the child did not return af- ter the previous investigation. On the other hand, it may be a new report. The case was troublesome, so the teacher decided to keep close watch on the child thus reported, and, if possible, keep it in school. In making Table XVI, however, a second report for the same period of absence was always considered as a separate report so that this would not account for the periods above three days. This plan nas been followed because it was not always possible to decide whether it was a new or a second report of the same period of ab- sence. The fact that suggestions only can be made to explain the periods reported more or less than three days, lends weight to the ar- gument that the records are not always properly and accurately kept on the attendance cards. Two cases will be given to illustrate the variation of days for which children are reported to the attendance officers. A boy eleven years old, in the second grade, was reported 15 times in one year. The sum of the days for which he was report- ed was sixty. This is an average of four days to a report. 12 times the attendance officer reported him as a truant. In another school, a girl fourteen years old, in the sixth grade, was reported n times in sixteen months. She was reported absent twelve and one half days. This is an average of only a fraction over a day for each report. A number of the reports were for a half day. The difficulty in explaining the above variations in the periods for which absences are reported has been increased by the fact that the teachers generally do not record on the attendance cards, in the space provided, the time of return of those reported to the attendance officers. The only way to determine from the attendance cards whether those reported have returned, is by means of the next report to the attendance officers. If the next report immediately follows the 46 previous one, it may be assumed that the child did not return after the former investigation. If there is no immediate report of ab- sence, the child is supposed to have returned to school after the at- tendance officer's visit to its home. The time of return of a reported child, however, should be recorded by the teacher in the space pro- vided on the attendance card. This would be useful information to the attendance officer in making further investigations of the same child. It would show him how promptly the parents of the child had placed it in school, after promising that it would return. The absence of the record of the return of reported children made it impossible to determine how promptly children return after the attendance officers have visited their homes to find out the cause of their absence from school. This information would be interesting and would also show, in part, the efficiency of the bureau that has charge of enforcing the compulsory attendance law or its ability or lack of ability to deal with this part of the problem that is confronting the bureau for a solution. The date of return, however, was recorded on some of the cards, but the smallness of the number and the doubt as to their accuracy would make the statistics obtained from such an investigation of such a character that they would be hardly repre- sentative of the conditions in the City or show the average time of return of those reported, and so would be misleading in any conclu- sions that might be drawn from them. Table XVI shows that there is not much difference between the boys and girls in the number of times reported. The boys were re- ported 485 times and the girls 517 times. The girls, however, sur- pass the boys in the higher number of days for which they were re- ported. In Table XVII is given the number of times each one was re- ported, by sex and grade. Since this investigation was limited to the reports of one year, difficulty was encountered in determining how many of the reported absences on the attendance cards were for the period February ist, 1908 to February ist, 1909. 3-26 in the column headed "Date Reported" on the attendance cards could stand for March 26th, 1908, or March 26th, 1907, or even a year or more be- fore the latter date. By comparing the date and grade, errors were eliminated, and a due degree of accuracy can be claimed for the table. The table shows that almost one-half of the pupils reported to the attendance officers were reported only once during the year in- 47 vestigated, and almost three-fourths are within the limits of two re- ports. There are still some difficult cases that claim the attention of the. attendance officers, as can easily be seen from the table. There is not much difference between the boys and the girls in the number of times for which they were reported during the year. There seems to be, judging from the above tables in this chapter, the same difficul- ty in keeping the girls in school as the boys. Taking into considera- tion the entire number of boys and girls in the schools investigated, their separate reports have been about equal in the per cents based on the original numbers. Table XVIII shows the reasons for absences as recorded by the attendance officers. The reasons have been classed under general heads. The statistics of the boys and the girls have not been re- corded separately. A few excuses, which could not be placed in any of the general classes in the table, have been listed under mis- cellaneous. This class contains also a few excuses that could not be definitely listed on account of the impossibility of deciphering the illegible writing. Sickness of the children or of members of their families account for over one-third of the reasons for absenteeism. The class marked "No investigation" needs explanation. Children were reported to the attendance officers, but in the space on the attendance cards provided for a report of the results of the investi- gation, no reasons were recorded. This class contains almost one- fifth of the cases reported and varied from eight to twenty-five per cent of all the cases reported, in the schools where this investigation was made. It hardly seems probable that a careful attendance officer would fail to record the results of his investigation. No direct state- ment was obtained to explain the failure to record reasons for ab- sence in the proper place on the attendance cards, but the require- ments of the law lead to the conclusion that "No investigation" ex- plains the lack of records. The "No reason" class represents the cases where the attendance officers recorded no reasons for the ab- sences but stated that action was being taken to place the children in school. It seemed advisable to make a separate list of the re- ports where there were records of action taken to place in school the children reported. In some cases a preliminary notice was sent when the reason recorded on the cards was sickness. Evidently the at- tendance officers had cause to doubt the excuses given to explain the reasons for the absences of the children under investigation. All the 48 TABLE XVIII. REASONS GIVEN FOR THE ABSENCE OF THOSE REPORTED TO THE ATTENDANCE OFFICERS, BY GRADE. EXCUSES & Second o 2 P ,0 2 1 E J3 "P c73 Seventh .5 J3 '-t : ~ 1 c 6 u 2 Sickness Sickness in family 6 20 11 53 13 52 7 74 31 68 13 13 7 4 1 290 83 28.94 8.28 Out 9 7 30 4 9 5 1 55 5.49 Out of City 2 2 8 4 11 4 31 3.10 Charity needed 33 19 1 9 2 54 5.39 Needed at home 3 4 6 18 13 5 48 4.79 Moved 2 3 1? 10 11 6 3 47 4.69 Truant 34 6 9 7 7 1 54 5.39 Miscellaneous 7 9 37 30 17 5 1 96 9.58 No investigation 33 77 48 31 43 13 193 19.26 No reason 8 4 10 18 8 2 1 51 5.09 Totals 8 142 156 198 237 196 56 9 1002 100.00 Action taken ~- Reported to Office 1 1 4 6 1 13 1.30 Preliminary notice 11 6 3 17 7 3 47 4.69 Interviewed 2 2 5 jj 3 1 1 15 1.49 Reported to charity ? 2 0.20 Totals 16 9 1? 3S 10 4 1 77 7.68 excuses classed under "No reasons" were inlcuded under the classes given in the list headed "Action taken". This list has made it possible to record double reasons, when found, and yet not make un- just comparisons between the different classes of reasons for ab- sences. Truancy was recorded as a reason for absence in very few cases. Since action was taken in all the cases classed under "No reasons,'' these can be added to the truancy class. This would make a total of over ten per cent of all reported classed as truants. It is not possible to determine from the attendance cards, to any great extent, the efficiency of the attendance officers. The number of cases where no reasons are recorded to show that investigations were made is an argument against their efficiency. The character of the excuses accepted by them in making their investigations, in part, determines their efficiency, but principals and teachers only can decide the extent of their efficiency in this matter. The manner in which the attendance officers conduct their investigations has an in- fluence on the children's return. It was not possible to test this. In Table XIX is given a number of troublesome cases that seem- ed to be of interest. The most important excuses have been given to show the character of the reasons given to the attendance officers. 49 TABLE XIX. A LIST OF TROUBLESOME CASES IN THE SCHOOLS INVESTI- GATED. BOYS GIRLS d <5-r e 8g ^ w d So .2 -g Most Important d So .2* Most Imprtant 2 cT! i^c Excuses be h fc Excuses O c ^ S O < et *, 2 9 12 15 Sickness, charity need- IT 8 13 6 Sickness ed. 2 9 8 12 Sickness, sent home 2 9 15 12 No shoes, no report, twice don't know 2 9 8 12 Sickness, truant, no 2 9 8 12 Sickness report 2 9 15 15 Sickness, charity, 2 10 7 6 Truant truancy 2 11 15 12 Truant (12 times), no- 3 9 9 12 Sick, out, no report tice and prosecuted 3 15 5 12 Needed, truant 3 10 11 12 Sickness, no report 3 15 6 12 Sick, sent to special 3 11 2 12 Moved, preliminary school notice 4 11 10 30 Sickness, no response 3 12 24 24 Sickness, no report, preliminary notice 4 11 16 16 No report (10 times) 3 13 10 4 Mother sick, charity needed, notice 5 11 7 12 Truant 4 10 4 6 Sickness, preliminary notice 5 11 2 12 Truant 4 12 10 16 Sickness 5 12 3 12 Truant 5 11 3 12 Sickness in family 5 12 * 1 12 Sickness, reported to 5 12 8 12 office 5 13 12 12 Truant 5 12 8 12 Sickness, moved, no good excuse 6 12 2 12 Truant, preliminary 5 13 6 12 Charity needed, pre- notice liminary notice 6 13 13 16 Sickness 6 13 12 24 Charity needed, sick- ness 6 15 2 12 Truant 6 14 3 12 Truant 6 15 3 12 Truant 6 14 11 16 Toothache, grippe, re- moved, truant 6 12 2 12 Truant, Preliminary 7 12 13 12 Mother sick notice To these statistics have been added the records of a few children who were not behind their respective classes in age. In these cases the reasons given are generally sickness, although a few cases of truancy were reported among these. Attention was paid to the truant class although all the truants have not been included in this table as it was not intended to make it complete but only representative of the difficulties of enforcing the compulsory attendance law. The results of the investigation show that the records on the attendance cards are not carefully kept. Important data are not 50 recorded. Investigations are not made in all the cases reported to the attendance officers. The law is not fully enforced in the case of the number of days absence for which children should be reported. The attendance cards are not always transferred" with the pupils. Attendance cards of those who have left school are often in the pos- a new attendance card was made out each year or every few years, session of the teachers. It was not possible to determine whether or the same card used for reporting until the child left school or the card was full. The cards seemed to reveal the fact that all the plans suggested were followed. In some cases there were two or three cards for the same child. Occasionally these cards were held by different teachers. The final destination of the cards could not be fully determined. The Bureau of Compulsory Education is under the impression that the attendance cards of all who have left school are sent to its office. Investigations favor the conclusion that all the attendance cards are not sent to the office of the Bureau. The attendance officers report troublesome cases to the Bureau of Compulsory Education. The past records of children thus report- ed are furnished. Investigations are made by the Bureau. The rec- ord of the trouble and its disposition is placed on the enrollment card of each child investigated, which cards are in the office of the Bureau of Compulsory Education. If the reported age of a child is doubted, the true age is determkied from the vital statistics of the city, the pass- port or baptismal certificate. Arrangements had been made to verify, in the same way, the age of those applying for employment affidavits. When the law of 1905, under which employment affidavits were issued, was declared unconstitutional, this plan could not be systematically carried out. Under the present law, a notary or magistrate can grant an employ- ment affidavit to any child whose parents declare, under oath, that the child is fourteen years old. This permits the child to obtain an employment affidavit without the knowledge of the principal of the school where he was last in attendance. If a notary or magistrate issues an employment affidavit to a child enrolled in school, the prin- cipal of the school reports the case to the Bureau of Compulsory Education, when the school record shows that the child is not fourteen years old. The Bureau verifies the age by means of the vital statistics of the city, passport or baptismal certificate. If the child is not four- teen years old, as shown by this verification, he is reported to the 51 attendance officer who tries to find him and replace him in school. The Bureau of Compulsory Education requires a statement from the principal of the school where the child attends, before granting an employment affidavit. The age is verified, and, if the child is found to be fourteen years old, the employment affidavit is granted. It can readily be seen that the present method of granting employment affida- vits has increased the difficulties attendant upon the enforcement of the compulsory attendance law in Philadelphia. It seems no more than reasonable that the same bureau which is charged with placing all the children in school and keeping them in regular daily attendance, so far as possible, should have the authority entirely to grant all the permissions to those who ask to leave school as soon as the law per- mits them. An investigation was made during the fall months of 1908 to find why children leave the public schools. This investigation cov- ered twenty schools of Philadelphia. The reasons have been sum- marized and are given in Table XX, classified according to age at time of withdrawal. Possibly most of the reasons for leaving the public schools, re- corded in Table XX, are legitimate. The table does show the diffi- culties encountered in enforcing the law of compulsory attendance. The number of children that left to enter the parochial school is very TABLE XX. REASON FOR CHILDREN LEAVING THE SCHOOLS, CLASSI- FIED ACCORDING TO AGE AT TIME OF WITHDRAWAL* REASONS OJ > x 4) | 1 3 | Z t Eleven Twelve $ fa 1 Sixteen 42 g To parochial schools . . Moved to unknown address Moved out of city .... Moved Absent 7 weeks name taken from roll .... Not returned since vacation Unknown 2 1 6 3 4 2. 7 2 9 7 5 7 1 1 9 3 11 7 1 1 13 9 8 4 1 1 7 8 13 2 1 6 1 4 1 3 1 7 4 9 1 3 1 9 3 9 1 3 1 3 2 2 1 2 1 1 61 42 68 31 17 8 8 Work . . . 7 t 17 52 11 ?, 84 Needed at home Non-attendance name dropped. 5 3 1 1 1 2 7 1 1 12 10 Sickness .... 1 1 1 1 ^ 3 1 1 12 Left for or placed in home Miscellaneous 1 3 6 5 4 3 1 9 1 1 5 4 ? 3 1 9 33 *The data in this table was furnished by a civic society of Philadelphia. 52 large. This varies for the different schools investigated. A large number go to work as soon as they reach the age of fourteen. Atten- tion should be called to the nineteen, under the work class, who have not yet reached the age limit which permits them to leave school and go to work. The moving of children from one district to an- other or out of the city presents difficulties for the attendance officers. In several cases the names of children were dropped for non-attend- ance. In the case of those dropped temporarily, it was often due to a legitimate reason for absence, as, for example, continued illness of the child. Some names were dropped entirely. It has not been possible to determine how many names are dropped from the school rolls each year and the reasons for dropping them. Table XX does not throw much light on the efforts made to replace the children in the public schools, or how far the attendance officers go in investi- gating the location of the children who leave the schools. CHAPTER VI. CONCLUSION. This investigation has been carried on with two ends in view; first, to see whether there has been an increase in the school enroll- ment and average attendance since the passage of the compulsory at- tendance law; second, to determine, so far as possible, what is being done to efficiently enforce the law of compulsory attendance. The first part of the investigation has been prefaced by a study of the influence of compulsory attendance laws on the enrollment and average attendance of the United States, its divisions and Pennsyl- vania. Every effort has been made to account for influences that may have a marked effect on enrollment and attendance. These influences have been noted and considered in making the comparisons and draw- ing conclusions therefrom. The results of the comparisons, howev- er, are somewhat disappointing, in view of the fact that the number of states under compulsory attendance laws is increasing. This is no argument that the laws of compulsory attendance are ineffective. Three explanations may be given for the decrease in school enrollment and average attendance during the years 1900 to 1907. This period showed a marked decrease for the United States, all its divisions, ex- cept the Western, Pennsylvania and Philadelphia. First, there may 53 be a less per cent of the population of school age, due to a lower birth rate and adult immigration. Investigation and comparisons of the last three census years favor the conclusion that the birth rate is steadily decreasing. The national census for 1910, however, will decide the question as to the influence of a lower birth rate and adult immigration on the school enrollment and average attendance for the period 1900 to 1910. Second, changed conditions may have their influences. Changes in the environment of children during the first part of the nineteenth century made necessary the passage of the truant law in 1850, and the compulsory attendance law in 1852, in Massachusetts. In fact, all the compulsory attendance laws may be said to be due to the changed environment of children. The formei methods and laws for placing children in school and keeping them in regular daily attendance, so far as possible, gradually became in- effective. There is no way, however, to determine what the school enrollment and average attendance would now be for any state or city under compulsory attendance laws, if such laws were not in force in said city or state. But a knowledge of the present condition of children, especially in the cities, and the commercial instinct that very often dominates them and their parents give some idea as to what the result would be in a city or state if the laws of compulsory attend- ance would be disannulled. No just comparisons can be made between cities or states with compulsory attendance laws and those without such laws, as the conditions would not be the same. Third, possibly the compulsory attendance laws are not efficiently and effectively efforced. This leads to the second part of the investigation where the test has been made to determine, so far as possible, whether the com- pulsory attendance law has been efficiently executed in Philadelphia. Four things seem necessary to enforce the compulsroy attend- ance law ; adequate school accommodations, a school census which en- rolls all the childern in the city or district, complete and accurate at- tendance records, and an efficiently organized bureau for enforcing the law of compulsory attendance. While the number of children to a teacher is decreasing for Philadelphia, as shown in Chapter III, the reports of children on half-time indicate that the present accommodations are not sufficient for all the children who apply for admission to the public schools or are brought in under the law. The further important question as to 54 whether school accommodations are adequate to the population of the respective districts was not touched upon by the investigation. The school census was considered, at length, in Chapter IV. The comparisons in that chapter seem to show conclusively that the school census of Philadelphia is too low. Either the children have not been found by the attendance officers in taking the census, or have escaped enrollment in the census by means of a false statement as to age. If a child or his parents inform the attendance officer that he is sixteen when he is only fourteen, or fourteen so that he is permitted to go to work, and informs the attendance officer that he has obtained an employment affidavit, when he is only twelve or thirteen, there is no way to check this false statement by the census. The school en- rollments must be used as a check in such cases. This requires that lists be made from the census by the Bureau of Compulsory Education. These lists must be sent to the schools and there compared with the enrollments. In this way there is a check placed on giving a false report of age to the attendance officers making the census, or, at least, such a false statement is detected and corrected so that the child or children will not benefit from the statement by non-attendance at school. This statement is made in the light of the great difference between the school and national censuses for Philadelphia, for 1900, as shown in Chapter IV, which increases as the work age is reached, where the difference is very great. If the Bureau of Compulsory Ed- ucation requires parents to declare on oath the age of their child to obtain an employment affidavit, and then verify this age by the vital statistics of the city, passport or baptismal certificate, caution should also be used in accepting the report of the age of children in making the census. This is especially needful if there is to be no check on the census further than comparing it with itself. The present method of comparing the school census made by the attendance officers with the school enrollment determined also by the same officers in making the census seems to open a fruitful field for deception by the parents and children and the eluding of the com- pulsory attendance law. This comparison is useful but should not be the only one used to check the census. In the light of experience at places where an effort has been made to test the accuracy of the school census, as described above in Chapter IV, checking and counter- checking would be suggested to test the school census of Philadelphia, although the checking hardly needs to be mentioned since the law 55 requires that to be done. The size of the city will not make the problem impossible or give added duties to the Bureau of Compulsory Education, that will seriously interfere with its present work. If the work be properly organized and divided, it can be done expeditiously and at the cost of little extra labor to anyone. This method of check- ing and counter-checking would compel the Bureau of Compulsory Education to obey the law requiring the making out of lists, from the school census, showing the names of all the children in each school district. These lists should be sent to the respective schools and a com- parsion iriade between the lists and school enrollments, under the sup- ervision of the principals. The names on the lists, which were found not to be on the enrollments, should then be added to the latter. These lists have been made out and compared with the school enrollments, in former years. The law requires that this be done and any failure to do it, as was the case in 1908, is contrary to the demands of the law. In making comparisons between the census lists and school en- rollments, a check mark should be placed on the enrollment cards. Enrollment cards could be made with spaces for such a check, yearly, sc that the check mark would not mutilate or disfigure the enroll- ment cards or interfere with a similar checking in other years. After the first comparison has been made and the names added to the school enrollments that were found on the lists and not on the school enrollments, the school enrollment cards would reveal how many names, by the absence of the check mark suggested above, were missed in making the census of the respctive school districts, of the children who came to school without the aid of the attendance officers' investi- gation. The names, thus found, should then be reported to the Bureau of Compulsory Education. After all eliminations had been made, if there were any due to the fact that some children had moved from one school district to another between the time of taking the cen- sus and making the comparisons, the remaining names could be ad- ded to the enrollment cards in the office of the Bureau of Compulsory Education. Of course, this plan would not furnish the names of those missed in making the census, who did not enroll in school in September of the year in which this comparison was made. If these names could also be obtained, the carefulness and integrity of the attendance officers in taking or making the census, could, to a great extent, be determined. In Chapter V the accuracy and completeness or lack of com- 56 pleteness of the attendance records have been- considered. Sufficient evidence has been given to show that the compulsory attendance rec- ords of the schools are not accurately kept. The cards contain suf- ficient columns to record all the necessary data resulting from a child's absence and the attendance officer's investigation of the cause for such absence. There should be, however, more care exercised in recording the data. No other records resulting from the enforcement of the compulsory attendance laws were investigated, except the printed re- ports of the Bureau of Compulsory Education. These reports, four in number, give scarcely any data to show what has been done to enforce the compulsory attendance law. The attendance officers are supposed to make regular reports to the Bureau of Compulsory Education, which give an account of all the work done by them. The printed reports of the Bureau, however, give very little information about the work of the attendance officers. This is especially noticeable when a comparison is made between the reports of the Bureau of Compul- sory Education of Philadelphia and similar reports of other cities w T here an effort has been made to give to the public as full an account as possible of the working of the compulsory attendance laws and what is being accomplished through them. The fact that no satisfactory evidence could be obtained as to the final destination of all the attendance cards adds weight to the state- ment that the records of the enforcement of the compulsory attend- ance law are not accurately and efficiently kept. It was also shown in Chapter V that the compulsory attendance law which requires that a child who has been absent for three days or their equivalent be reported to the attendance officer, was not always obeyed by the teachers in reporting children who had been ab- sent. Two solutions might be given to explain this. Either the teach- ers are ignorant of the demands of the compulsory attendance law with respect to the period for which children who have been absent should be reported, and how to determine the length of the period, or they have no regard for the requirements of the law and prefer their own system of reporting absentees. It is not possible to decide whether the excuses recorded in Ta- ble XVIII were lawful or not. The immediate return to school of 3. child reported absent is the end sought. The attendance officers must realize the fact that their efficiency is determined largely by the number of children they place in school, both of those reported to 57 them as absent and those they find on the streets of their respective districts, who are of the compulsory school age. Excuses are useful if they help to place the children in school. The amount and char- acter of the instruction, exclusive of any printed regulations, that is given to the attendance officers by the Bureau of Compulsory Educa- tion, has not been determined. From some of the reports received from principals and judging from the reports on the attendance cards, it would be recommended that the Bureau give instructions to the attendance officers in order that they may properly and efficiently per- form their part in the enforcement of the compulsory attendance law and not degrade their position to one of obtaining reasons for absences. It has been shown that the compulsory attendance law does not properly provide for the granting of employment affidavits. If the Bureau of Compulsory Education is charged with the responsibility of placing in school, and keeping there in regular daily attendance, all the children between 8 and 14 years and those not employed between 14 and 16 years, it should be given the power to grant all the employment affidavits. Whether the Bureau of Complsory Education has an efficiently organized system for enforcing the compulsory attendance law can, in part, be decided from the facts given above. The lack of an accurate school census and incomplete and carelessly kept attendance records seem to show that the present organization of the bureau for enforc- ing the compulsory attendance law is not sufficient or complete enough to meet the demands made upon it or properly regulate or use the forces that must co-operate with it. The blanks used in making the different reports used in enforcing the law, so far as they have been examined, seem to be adequate for the purpose. The lack of unifor- mity in the school records which are used in enforcing the compulsory attendance law, and the carelessly recorded data on the attendance cards have the marks of inefficiency. The cause and responsibility for this state of affairs will not be centered on the Bureau of Com- pulsory Education entirely since it is not known what authority the Bureau has in making demands of the teachers to furnish the data necessary for the enforcement of the compulsory attendance law, in the way that it may deem most advantageous for such enforcement. The Bureau of Compulsory Education, however, realizes that the records on the attendance cards are not kept as accurately as they should be. It is not known whether any steps have been taken with 58 a view to correcting this deficiency and incompleteness of the attend- ance records in the schools. Inadequacy of records, looseness of organization and disregard of certain provisions of the law have been clearly proved. In the ab- sence of adequate records, the operation of the compulsory attendance law can not be accurately tested. The absence of adequate records is in itself a demonstration of inefficiency and lack of organization. Under such conditions, the law of compulsory attendance can not be efficiently enforced. 59 CHIEF SOURCES. Annual Report of the Board of Education of Philadelphia, 1877-1907. Annual Reports of the Superintendent of the Public Schools of Phil- adelphia, 18974908. Reports of the Bureau of Compulsory Education of Philadelphia, 1902, 1904, 1905, 1908. Annual Reports of the Superintendent of Public Instruction for Penn- sylvania, 18774908. Annual Reports of the United States Commissioner of Education. Journal of the Philadelphia Board of Education, 1892-1908. Annual Report of the Superintendent of Public Schools of Fort Wayne, Indiana, 1907. Annual Report of the Board of Education of Boston, Massachusetts, 1907. Annual Report of the Board of Education of Buffalo, New York, 1906. Annual Report of the Board of Education of Cleveland, Ohio, 1907. Annual Report of the Superintendent of Public Schools of Harristourg, Pennsylvania, 1907. Annual Report of the Board of Education of New York, 1907. Annual Report of the Board of Education of Chicago, Illinois, 1907. United States Census Reports, 1870, 1880, 1890, 1900. United States Census Yearly Reports, 1905, 1906, 1907. AUTHORITIES. F. H. Giddings, The School and Legal Aspect of Compulsory Educa- tion and Child Labor, National Educational Association, 1905 Re- iport, pages 111413. G. H. Martin, Child Labor and Compulsory Education, The School As- pect, National Educational Association Proceedings, 1905, pages 103411. James D. Sullivan, A Summary of the Compulsory Attendance and Child Labor Laws of the States and Territories of the United States, Albany, New York, State Educational Department, 1907. F. V. N. Painter, Luther on Education. The Early Withdrawal of Children from School, Addresses delivered 'before the Philadelphia Teachers' Association, 1907. L. Von Bokkelen, Compulsory Education, ^Circular of information of the United States Bureau of Education, December 1871. George H. Martin, Evolution of the Massachusetts Public School Sys- tem, 1894., Wm. H. Maxwell, The Nation, April 25th, 1907, page 380. Wm. H. Maxwell, Present Problems of the School, Educational Re- view, November 1904, V. 28, pages 378-395. Julia Richman, Tlhe Incorrigible Child, Educational Review, May 1906, V. 31, pages 484-506. Stratton D. Brooks, Causes of Withdrawal from School, Educational Review, Novemlber 1903, V. 26, pages 362-393. J. W. Perrin, Beginnings in Compulsory Education, Educational Re- view, March 1903, V. 25, pages 240-248. Indirect Compulsory Education. The Factory Laws of Massachusetts and Connecticut, Educational Review, April 1906, V. 31, 383-394. J. W. Perrin, History of Compulsory Education in New England (The- ta). Snedden and Allen, School Reports and School Efficiency. DAY AND OVERDUE. 1933 UNIVERSITY OP CALIFORNIA LIBRARY