J^it'-^A ill pi TEXAS STATE BOARD OF HEALTH A STUDY OF THE SCHOLASTIC CENSUS AS TO ITS ACCURACY AS THE BASIS OF THE FREE SCHOOL FUND W. A. DAVIS, M. D. State Registrar of Vital Statistics BUREAU OF VITAL STATISTICS AUSTIN, TEXAS VON B0ECKUAMN-J0NS8 CO., FBINTEBS, AUSTIN, TEXAS A225-218-1M TEXAS STATE BOARD OF HEALTH A STUDY OF THE SCHOLASTIC CENSUS AS TO ITS ACCURACY AS THE BASIS OF THE FREE SCHOOL FUND W. A. DAVIS, M. D. State Registrar of Vital Statistics BUREAU OF VITAL STATISTICS AUSTIN, TEXAS TON BOECEMANN-JONES CO., PRINTEBS, AUSTIN, TEXAS / D. of D. APR 29 1918 A STUDY OF THE SCHOLASTIC CENSUS In 1917, the State Board of Education divided $9,241,643.50 among the public schools of Texas. This, division was based on the scholastic census taken in April of the same year, which enrolled 1,832,219 chil- dren within the scholastic age, and resulted in a pro rata of $7.50 per pupil. From the school funds of the State each child from 7 to 17 years of age is entitled to this amount and each community draws $750 for each 100 scholastics. There is, under the present law, no other method of making the appropriation for each school. If the census was correctly taken, then the pro ]-ata is just and each child gets its propor- tion of the public funds. If the census was not correctly enumerated, then some child is robl)cd of its part of the funds, and the pro rata is not just. If any community has enrolled more children than are in fact within the scholastic age, then that community is receiving more than its share of the public funds, and the pro rata is not justly ap- portioned. So far as is known to the writer, no fraud has been practiced, no official has padded the census rolls, nor has any official been negligent of his duty: the scholastic census is not sufficiently accurate to be made the basis of such an appropriation, nor does it safeguard the interests of the children and the interests of the State. The accuracy of any census depends directly upon the honesty, the intelligence, and the precision of the one who furnishes the data taken in the enumeration. Very apparent mistakes appear in all census re- ports. With all the safeguards thrown about the Federal Census, there are from time to time irregiilarities that have become regular. For ex- ample, such reports would appear to show that more persons are of even than of odd numbers of years of a.ge, which would be extraordinary. It is hardly possible that more people die at an age ending in or 5 than at an age ending in 1, 2, 3, 4, but such would ap]3ear to be a fact from census reports. And, so long as the enumerator is dependent upon the careless and uneducated, who see so little value in correct cen- sus taking, such irregularities will occur. In the enumeration of the scholastic census several motives may play some part in misrepresent- ing the exact age of the child. The parent may wish to hold the child out of school so as to secure the benefits of its labor; or, on the other hand, not wishing to lie bothered with the child at home, may raise its age so as to get the benefit of the free school as a kindergarten. It may be to the advantage of a community to enroll a greater number of chil- dren, and in this way secure more than a just amount of the public money. And, while the enumeration of the children in each family must be sworn to by the parent or guardian it is apparently an easy matter for mistakes tn enter into the enumeration to such an extent that the Compulsory School Attendance Law may be rendered to a cer- tain extent null and void and the child's chances to secure an education may be lost. It would appear that when two enumerations are taken for the same territory, for the same ages, and for the same year that the results should be the same. In 1910, the Federal Census shows 928,805 chil- dren between 7 and 16 in Texas; while the State scholastic census shows for the same year and ages 968,269, a difference of 39,464 children. Both enumerations were taken in the early paii; of the year, by local men paid per capita, and while one or the other might have overlooked some children within the age, it would be raaiiifestly unjust to say that the Federal enumerators were more careless than the scholastic enumer- ators to the extent of overlooking 39,000. The pro rata for that year was $6.50, and the justice in appropriating $255,248 for the education of this 39,000 may be questioned. The only difference that could enter into the two enumerations has been mentioned above, and if that be the case such errors cannot be corrected under the present system. The 1910 scholastic census shows that 113,879 children entered the public school age; for the same year and same age the Federal Ceiisus show only 105,173, a difference of 8706; while at 16 years of age the former shows 79,937 and the latter 88,293 children, a difference of 8356. The difference shows that the scholastic enumerators enrolled 8706 more at the age of 7 and 8356 less at the age of 16 than was en- rolled by the Federa.l Census enumerators. Such enrollment placed 8706 children at 7 years of age in school and took from school 8356 children at 16 years of age, contrary to the enrollment of the Federal Census. The cut opposite this page illustrates the irregmlarities that are to be found in the census. On the left margin is scaled the number of children; on the lower margin the age of enrollment. It will be noted that, according to the 1916 scholastic census, 138,000 children entered the public school age, while 68,000 children passed out at the age of 17. The 1916 census gives 36,000 more children at the age of 7 than is given by the 1917 census; while the 1916 census enrolled 35,000 less at the age of 17 than were enrolled by the 1917 census. The Federal census of 1910. as well as the scholastic census of 1910, 1915, and 1916, show a greater numlDcr entering the school age than leaving that age; the reverse is true as to the 1917 census, which reports 102,000 enter- ing and 104,000 leaving the school age. A careful study of those ir- regularities as shown in this chart will develop many such apparent errors. It is probable that the Federal Census is not absolutely accurate, but the irregularities of the Texas scholastic census are very noticeable. By reference to Table I, it will be noted that the increase by years for the seven years following the 1910 census is extremely irregular. The in- crease as shown is as follows: During 1910 an increase of 1372; dur- ing 1911. 1888; during 1912, 3426; during 1913, 4782; during 1914, 2646: during 1915, 10,765: while during the year 1916 there was a de- crease of 36,348 children at the age of 7 years, as shown by the 1917 census. If the number of children enumerated as 7 years of age in 1917 is correct, then it is hardly possible that the number shown for that age in the previous years can he correct, since there was no partic- ular emigration during the year 1916. — 6— TABLE I. The Federal Census and the Scholastic Census for Texas by Years of Aee From 7 to 17. Federal Census. Texas Scho'astie Census. 1910 1910 1911 1912 1913 1914 1915 1916 1917 7 105,173 100,699 93,373 94,875 113,879 108,914 101,783 Group A 99,859 115,251 111,686 105,524 105,232 89,062 117,139 114,430 108,104 107,881 100,361 98.866 120,565 117,892 111,284 110,946 104,568 100,701 101,138 125,347 122,695 116,071 115,736 109,355 105,493 105,927 104,526 127,993 123,955 118,578 116,359 111,239 110,105 108,626 107,145 104,118 138,758 130,933 121,139 121,731 113,826 116,818 110,594 106,695 100,591 93,190 102,410 8 122,556 9 113,206 10 116,802 11 84,676 Group B 91,880 111 910 12 97,336 87,795 92,110 84,475 88,293 98,855 93,669 93,102 86,391 79,937 96,545 117,701 13 95,870 94,135 8S,173 80,931 98.110 113,603 14 96,697 91,859 83,779 99,739 115,965 15 95,794 85,943 100,588 113,838 16 90,729 100,934 109,826 17 68,944 104,402 Total 928,805 96«,269 991,409 1,017,133 1,049,570 1,096,467 1,129,152 1,223,219 1,232.219 The increa,se of 10,765 for 1915, as shown in the census for 1916^ when compared with the increase of 2646 for the year previous, cannot be accounted for in any scientific manner. The children 7 years old in 1916 were bom in 1909. 1907 was marked by the greatest panic in the history of the State. The effect of this panic which occurred in the fall was especially felt in the year 1908, and effected the birth rate of 1909. A high birth rate is always an indication of prosperity and the reverse is generally true. 1909 would show a lowered birth rate and the scholastic census should show a decrease of children born that year when compared with other years rather than an increase of from two to seven times greater than that of previous years. If the scholastic census is correct, the enrollment of only 102,4^10' children at 7 years in 1917 would indicate that in 1917 there were 11,469 less than in 1910, a decrease rather than an increase during the seven years from 1910 to 1917. If the price of wheat and the value of the cotton crop be considered in connection with the scholastic census, it will appear that the num- ber entering the public schools is not controlled by the number of chil- dren but by the economic condition of the State. By reference to Table 2, it will be noted that there was no great difference in the value of the cotton crop during the years 1910, 1911, 1912, and 1913, if the price be considered in conjunction with the cost of picking and ginning, but in 1914 when the European war had closed the cotton exchanges, and there was no market, the value of the crop dropped off $79,570,000 in value, not including the cost to the farmer of picking and ginning a four-million-bale crop, at the same time wheat advanced 18 cents per bushel, the scholastic census of 1915 shows an increase of only 2646 at the age of 7, as compared with 4782, the increa.5e of the year before. — 7— In 1915, the cotton crop tiliows a shortage of 1,322,000 bales, with an increase of more than $22,870,000 in vahie. This shortage saved the farmer more than $15,000,000 in picking. "Wheat had declined 6 cents per bnsliel in price, making flour ehoapor and the 1916 census shows the unusual increase of 10.765 children entering the scholastic age, or a total increase for all ages of 94,067 compared with 32,685 of the year before. TABLE II. Cotton Crop. Year. Price Per Pound. Number of Bales. Total Value of Crop. Wheat Per Bushel. 1910 14.0 9.6 11.5 12.5 7.3 11.2 2,949,000 4,107,000 4.645,000 3,773,000 4,390,000 3,068,000 $ 241,310,000 238,170,000 312,610,000 276,600,000 197,030,000 219,900.000 88.3 1911 87 4 1912 76 1913 79.9 1914 98.6 1915 92.0 Increase and Decrease in the Value of Cotton and the Price of Wheat Over the Year Previous. Year. Value of the Cotton Crop. Price of Wheat. Increase in Scholastics. 1911 Decrease $ 3,140,000 Increase 74,440,000 Decrease 36,010,000 Decrease 79,570,000 Increase 22,870,000 Decrease Decrease Increase Increase Decrease .9 cents 11.4 cents 3.9 cents 18.7 cents 6.6 cents Increase 23,140 1912 Increase 25 724 1913 1914 Increase 46,897 1915 Increase 32 685 1916* Increase 94,067 1917* Increase 9,000 ♦Statistics of cotton and wheat crop not compiled for 1916 and 1917. It must be remembered that each million bales of cotton cost at least 513,000,000 to pick and gin and that while 1911 and 1914 show a 4,000,000 bale crop that this cotton at 9.6 and 7.3 cents per pound was gathered at a loss to the farmer. The census of the next year will show this efi'cct by proportionately lower increase in the scholastic census. This effect is very apparent in the increase of scholastics in 1915 census as influenced by the conditions of 1914, and the prosperity of 1915 resulted in the greatest increase shown in any census report. The unsettled condition resulting from the increased price of food and merchandise and the declaration of war in April when the 1917 census was taken caused a most marked decrease as shown by the loss of 85,000 in the scholastics as compared with the increase the year before. It is possible to account for the increase in those entering the scho- lastic age by taking into consideration the element of immigration, but the decrease cannot be accounted for by emigTatiou. While it has been suggested that the transient Mexican population causes this extreme variation, close study of that clement will contra- dict the statement. In 1915, the total number of Mexican children enumerated was slightly over 104,000, while in 1916 the total was 105,- 000, an increase of only 1000. If a close comparison he made, it Avill be found that there is less variation in the number of Mexican children enumerated than among other nationalities. It is a well known fact that some communities may decrease, but on the whole the jwpulation of the State i.s. rapidly increasing; so rapidly in fact the usual methods of estimating the population during the in- tracensal years are not reliable, when applied to Texas. The inaccuracies in the scholastic census i.s apparent if any particu- lar group of children be followed tlirough the school ages. Take, for example. Group B, Table I. 91.88U ch'ldren were 11 years of age in — 8— 1910. Tliese children were 12 years of age in 1911, and had increased from 91,880 to 96,545 in number; the same group of children increased in 1912 to 98,110 and in 1913 to 99,739 ;"tlie next year to 100,588, and in 1915 to 100,934, making a total increase of 9054 during the five years previous; running from 33fi to 4655 increase per year. The ques- tion arises, where did this number of children come from? Immigra- tion is the only reason that can be given for this increase. The scho- lastic census shows that this 91,880 children which had increased to 100,934 in five years dropped off in 1916 to 68,944, a loss of 31,990 during the year 1915 or 22,936 during the five-year period. Such a loss cannot be ascribed to emigration, and must be due either to death or to the methods of taking the scholastic census. If Group A, Table I, be followed through the school ages, it will be found that the 1911 census shows a decrease of 10,797 children in this group in the year 1910, while the 1912 census shows this same group to have increased by 9804; the 1913 census shows an increase of 2272; the 1914 census an increase of 3388; the 1915 census a decrease of 408; and the 1916, a decrease of 10,928; and the 1917 census an in- crease of 11,212. It is impossible that the population should fluctuate to this extent, and such variations must be ascribed to the lack of ac- curacy in the census. While the 10-vear-old group in 1910 decreased 10,797, the 11-year-old group, the same year, had increased 4665; the next year the one group increased 9000, the other increased only 2000. Such comparison may be made as to each year with the result that there is no similarity in the increase or decrease; that there is no law controlling the variation. By reference to Table HI, the death rate of children in the scholastic age will be found and the decrease in any year will in every case be greater than the death rate. TABLE III. Deaths by Ages and Years of Children Within the Scholastic Age Computed on Rates Taken From the Federal Census Life Tahles. Death rate Age. per 1,000 1910 1911 1912 1913 1914 1915 1916 Total Deaths by Ages. 7 3.30 37.5 380 314 386 322 267 397 332 274 251 413 345 286 262 422 349 293 264 243 244 457 369 299 276 249 259 261 2,830 8 2.82 307 2.51 Group A 226 Group B 201 2,338 9 2.47 260 238 195 1,930 10 2.27 244 219 219 1,761 11 2.19 229 223 238 239 1,575 12 2.22 219 221 239 245 252 214 234 249 268 1.612 13 2..36 226 241 250 255 231 256 275 295 1,682 14 2 57 248 • 260 264 256 274 285 294 1.801 1951 1.5 2.84 272 271 285 1.892 16 3.16 286 318 1,940 Total deaths 7 to 16, inc 2,536 2,573 2,660 2,743 2,867 2.959 1,.'^05 3 023 1,735 19,361 — 9— Of the 99,So9 enrolled in 1910 at 10, 1441 died within the next live years, leaving 98.418 at the age of IG. Of this nuniher, 294 died dur- ing 191G, leaving 98,124 in 1917, as eoin]);ired with the 104,402 on- rolled, if no emigi'ation oecnrrcd. Of the 91,880 in 1910 at 11 years of age, 1505 died, leaving 90,375 to be enrolled in 1916, rather than 68,94-1, if no emigration occurred. If, for example, the children 12 years of age in 1910 1)e taken, there appears a loss of 8126, as shown by 1914 census, when, according to the death rate for the ages included, the total number of deaths would be only 1251 children, leaving 6875 children who had been enrolled in the school age during 1910, 1911, 1912, and 1913, who were not enrolled in 1914. A detailed study of the scholastic census will develop many such errors. Any group if followed through the different ages will show variations that are not to be explained in any manner otlier than by an error in census emuneration, and the question presents itself, is such a census, honey-combed with errors, sufficiently reliable to be made the basis of the division of more than $9,000,000, and should the education of the masses be placed upon such an uncertain foundation? Should the parents under the Compulsory School Attendance Act be allowed to change the age of the child at their pleasure, and in this Avay deprive the child of its education and thwart the purpose of the founders of our public school system. If justice would be done to all, the age of the child must be made a matter of State record, so that it 7nay not be changed at any time or by any one ; and from this record should be taken a list of the children who are within the school age, and upon this list should be based the division of the public school funds. So long as it is the duty of the enumerator to enroll all children within the school ages, regardless of residence, whether permanent or transient, just so long will some dis- tricts receive more of the public funds than they are justly entitled to So long as the parent is allowed to give each year the age of the child, just that long Avill the economic conditions effect the number enumer- ated, and the children of those who are careless, ignorant, or dishonest may be deprived of an education. If the Com]3usory School Attend- ance Law is to be enforced, if the very children that should be educated at the expense of the State, and for whom the public school is especi- ally intended, are to l)e protected and are to receive an unintemipted education, the apportionment of the public funds and the enrollment of the child I'en must lie jdaced upon a more reliable basis than tlie scho- lastic census. LIBRARY OF CONGRESS 029 487 038 6