Popular Science Monthly/Volume 63/June 1903/Changes in the Age of College Graduation




THE belief seems to have become general that the American boy of to-day takes his first collegiate degree—A.B. or its equivalent—a good deal older than his father took his, and a great deal older than his grandfather. The present study was undertaken with a view to determining from actual records the measure and rate, if real, of this increase. The plates and tables that are presented herewith tell, in the main, their own story; my task will be little more than the making of a running commentary upon these.

The calculations are based upon nearly twenty thousand cases, and include the graduates of eleven colleges, representing all parts of the country except the extreme west. If undue weight seems to be given to the New England colleges, my excuse is twofold: first, the proportion of colleges that date back fifty years or more is much larger in New England than elsewhere; secondly, I have used all the published material I have been able to find, in the shape of alumni catalogues which give the date of birth of graduates. These have, moreover, been largely supplemented by private information very kindly furnished by the officers of colleges whose general catalogues do not come down to the year 1900.

The results are given in decade periods for the double reason that shorter periods are unwieldy, becoming too numerous, and because the longer period is more reliable. Two-or three-year periods often show what seems a very decided trend in a given direction; but this is in all cases decidedly modified if not entirely obliterated by the addition of the remaining years of the ten. The results thus win stability and evenness.

Before beginning the discussion of the tables and plates, one further word of explanation may be given. It will be noted that in Table I. and elsewhere the median age is used rather than the average age. The reasons for using the median age—the point above which and below which, respectively, one half of the students in each decade graduate—are evident. In the first place, the labor of finding the exact arithmetical average of the age of graduation of 20,000 students would be enormous; and when found it would not give us what we wish, viz., the age at which the students, or a definite percentage of them, actually do graduate. It is evident that a few students graduating in a class above forty years of age—by no means an unheard-of state of affairs—would unfairly raise the average age of that class, since it is manifestly impossible to graduate twenty years below the normal age. Again, a class, or series of classes, may graduate a considerable number of its members below twenty, while a still larger number graduates above twenty-four or twenty-five. The curve of distribution of the ages of graduation will then resemble the letter M. Manifestly, in such a case, which occurs several times, the arithmetical average tells us nothing of value. Finally, the median age gives us the exact information that one half the students in question graduated at or above the given age, and the other half at or below it. The curves of distribution, moreover, given in the plates for all graduates and all colleges for the years 1850-59 and 1890-99, show exactly what percentage graduated at each age.

Table I.

Median Ages of Graduation by Decades.

Dartmouth. Bowdein. Middlebury. U. of Ver. Adelbert.
Age. No. Age. No. Age. No. Age. No. Age. No.
1770-79 23-0 78
1780-99 23-1 150
1790-99 23-2 336
1800-09 22-6 323 22-10 76
1810-19 22-9 330 23-1 194 20-4 106
1820-29 23-1 328 23-0 187 20-8 258 22-4 59
1830-39 22-5 384 23-4 242 21-7 289 22-7 80 23-0 41
1840-49 23-1 586 22-8 109 21-9 356 22-0 184 23-2 125
1850-59 23-8 558 23-3 121 22-1 335 22-4 168 23-0 98
1860-69 23-1 491 23-5 132 22-10 348 22-6 91 22-10 160
1870-79 22-10 593 23-4 111 22-5 321 22-6 98 22-9 217
1880-89 22-10 527 22-11 86 22-8 303 22-8 108 23-0 251
1890-99 22-9 678 23-2 125 22-7 481 22-9 215 22-9 156

We now come to a consideration of Table I.[1] The most obvious and surprising thing that strikes us at first sight is the fact that our assumed great increase in the age of graduation, taken generally and so far as our material reaches, is absolutely non-existent.

The median age of graduation in Dartmouth, for instance, has in one hundred and thirty years fallen three months; in one hundred years the median for Middlebury has risen four months. But note that in 1830-39 the median for Middlebury was two months higher than now. In the case of Bowdoin, there has been a steady rise to a little over two years, which, however, reached its maximum in the decade beginning in 1860, and has since been falling. In seventy years, the University of Vermont median age has risen but two months; while in the same period that of Adelbert College has fallen three months. Again, we may compare the New York University with Oberlin College. While the age at the former has in sixty years risen one year and five months, in the latter it has fallen one year and seven months. It may be noted in passing that the number of graduates in the given

Table I.

Median Ages of Graduation by Decades.

U. of Ala. N. Y. Uni. Wesleyan. Oberlin. De Pauw. Syracuse.
Age. No. Age. No. Age. No. Age. No. Age. No. Age. No.
20-4 57 20-2 73 23-0 107 24-11 34
20-3 126 20-3 147 23-3 231 25-6 122 21-7 63
20-9 173 20-7 102 23-4 231 25-2 120 22-9 89 23-11 28
20-0 48 20-8 128 24-0 260 24-0 176 23-2 115 24-0 29
20-3 66 21-6 141 23-8 325 24-3 270 23-1 230 24-6 138
20-0 209 21-1 154 23-3 323 24-3 267 23-2 317 23-9 224
20-2 270 21-8 115 23-6 456 23-11 403 23-9 371 23-11 264

time is in Oberlin about double that in the New York University. Finally, we may call attention to the fact that in the University of Alabama and in Syracuse University, the age of graduation has remained practically unchanged, with a slight tendency to decrease.

So much for the general aspects of Table I. It will be of some interest to consider somewhat closely the changes that have come within the last two generations of college graduates, or since 1850. At this period all the colleges in our list are available for comparison; and it is since the beginning of this period that practically all the modern development of the American college has taken place. What happened before 1850, while it may be interesting, can not have the importance for us now that the changes of the past fifty years have.

At the outset, we note that of the eleven colleges in the table, the median age for one only remains quite unchanged—Syracuse. The following show increases, in months: Bowdoin, 6; Vermont, 5; New York University, 13; Wesleyan, 2; DePauw, 12; total, 38. The if following show decreases, thus: Dartmouth, 11; Adelbert, 3; Alabama, 7; Oberlin, 15; Middlebury, 1; total, 37.

Table II.

Average of Median Age of Graduation foe Past Fifty Years.

1850-59 1860-69 1870-79 1880-89 1890-99
Dartmouth 23-8 23-1 22-10 22-10 22-9
Middlebury 23-3 23-5 23-4 22-11 23-2
Bowdoin 22-1 22-10 22-5 22-8 22-7
Univ. of Vt. 22-4 22-6 22-6 22-8 22-9
Adelbert 23-0 22-10 22-9 23-0 22-9
Univ. of Ala. 20-9 20-0 20-3 20-0 20-2
N. Y. Univ. 20-7 20-8 21-6 21-1 21-8
Wesleyan 23-4 24-0 23-8 23-3 23-6
Oberlin 25-2 24-0 24-3 24-3 23-11
DePauw 22-9 23-2 23-1 23-2 23-9
Syracuse 23-11 24-0 24-6 23-9 23-11
Av. of Totals 22-9.6 22-9.3 22-9.9 22-8.3 22-7.5

The net result of the changes that have come in the age of graduation in these fifty years is more clearly presented to the eye by Table II. Here is presented a view of the medians for all the eleven colleges, wherein each college is given an equal weight, regardless of whether it be a large or a small college. By this method then is avoided the overweighting which a large college, like Dartmouth or Bowdoin, would otherwise exert on the results. The results show that in only one decade is the average of medians as high as that of 1850-59. Moreover, the last two decades show a slight decreasing tendency, making a net reduction in fifty years of two months for all the colleges.

Thus far we have dealt with the median age of graduation as distinct from the average age, and reasons have been adduced to show why the former is preferable to the latter as the measure in our present study. Inasmuch, however, as the arithmetical mean is the one in most common use, and further, as some may still feel that it, if investigated, would show the rise that has been supposed to exist, we will consider the data and results that Table III. shows. In this table are shown

Table III.

Average Age of Graduation for the Past Fifty Years.

1850-59 1860-69 1870-79 1880-89 1890-99 Cases.
Dartmouth 23-9.4 23-6.7 23-4.9 23-1.3 23-2.7 5362
Middlebury 23-8.1 23-6.5 23-5.8 23-6.5 23-8.1 1386
Bowdoin 22-6.4 22-11.7 23-0.0 23-1.6 23-2.4 2797
Vermont 22-11.5 23-3.3 22-8.6 23-3.4 23-0.2 1003
Adelbert 23-9.6 23-7.2 23-2.4 23-2.4 22-10.8 1048
U. of Ala. 21-0.0 20-1.8 20-2.4 20-3.6 20-6.0 949
N. Y. U. 21-1.6 21-2.3 20-8.4 21-7.5 21-10.8 860
Wesleyan 23-10.8 24-3.3 24-2.8 23-10.2 23-6.1 1933
Oberlin 25-0.7 24-7.5 24-5.3 24-8.7 24-3.9 1392
DePauw 22-2.4 23-8.4 23-8.4 23-9.1 23-10.3 1185
Syracuse 24-1.6 24-5.0 24-7.7 24-8.6 24-7.5 755
Av. of Totals 23-1.3 23-3.4 23-0.8 23-2.3 23-1.9

the arithmetical averages of each college by decades, supposing that the students graduating at any given year of age, say 32, are about equally distributed throughout the months of the year, thus giving an average for the given year of, say 22.5 years. With small numbers,

PSM V63 D167 Age difference of graduates within two generations.png

this assumption is not without its liability to error; but with numbers so large as we have, the errors are found by actual trial practically to negative each other; so that we can rely upon the results as being, for all practical purposes, and in the main, substantially correct.

The first striking thing to be observed in Table III. is the fact that the average age is a few months higher than the median throughout in the totals of all colleges. In the past fifty years, the average age of graduation has remained quite unchanged, while in the past forty years, the average has fallen one and a half months. This difference

PSM V63 D168 Age of college graduation 1.png

is, however, probably too small to be in itself significant, so that we may conclude that there is neither any actual change in the average, nor any definite tendency observable towards rising or falling.

In the above discussion of averages, each college has been given the same weight as every other. Now, we may look. at the same matter from another point of view. We may bunch all the graduates, as though

PSM V63 D168 Age of college graduation 2.png

they were all students of one great college; and, still assuming that they will be about equally distributed through the months of any given year—an assumption which by the now very much larger numbers is made doubly secure—we may take the average for the five decades since 1850. By this method we obtain the following results:

1850-59. 1860-69. 1870-79. 1880-89. 1890-99. 1900- .
Yr.M. Yr.M. Yr.M. Yr.M. Yr.M. Yr.M.
Av. 233.0 235.4 234.8 233.9 236.1 30.5

Even here, where every concession possible is allowed to the weighting of the averages by the few colleges which in the last decade have relatively much larger numbers, together with their consistently higher average age of graduation than in the earlier decades, we still find no change of any significance. At the very best, or worst, the change in fifty years past has been only three months. While now, if we may use for the sake of further illustration the available data of the colleges for the decade beginning 1900, we find on an average three months less than that of 1850-59. The colleges included here are those seven which furnished for the decade 1890-99 over 81 per cent, of all graduates, and include all the colleges except New York University, Adelbert College, Middlebury College and Syracuse University.

PSM V63 D169 Age of college graduation.png

It will be noted that all the largest colleges are included, and that of those omitted two are above and two below the average in the decade 1890-99.

We may now turn from the consideration of the tables to an examination of the plates. Plate I. shows the percentage of students actually graduating at each age—16 years to 31 years—in which last category are bunched for convenience all graduates of the age of 31 years or over—for the two decades 1850-59 and 1890-99, respectively. The upright line on the base in the twenty-second year marks the actual median age of graduation of all students for the decade. It will be noticed that its position remains absolutely unchanged. Perhaps the most noticeable exhibition presented by this plate is the pushing of the great bulk of graduates in the last decade into the comparatively narrow compass of the years 20-24, and the consequent great reduction of the numbers graduating above or below these limits as compared with the earlier decade.

One further observation is worth making: At first sight it appears that the mode—the year in which the largest number graduates—is in

PSM V63 D170 Age of college graduation.png

the first decade, the twenty-first year; while in the second decade this has been pushed up, and is now the twenty-second. In this there are two matters of significance. First, while the mode in the first decade is 21, the percentage here is still less than it is in the same year in the next decade, where the mode appears as 22; secondly, the reduction of the percentages in the years below the twenty-second in the second decade is largely due to the fact that in the first decade two or three colleges which have a high median age of graduation have in this decade very few students, while in the last decade they have a relatively very much higher number of graduates, thus acquiring an undue influence in the second decade, and failing to exert this influence in the first

PSM V63 D171 Age of college graduation.png

decade. This fact, which does not come out in this plate, becomes much clearer if we take decade 1860-69 for comparison with decade 1890-99.

Plates IV., V., VI. and VII. present the evolution of the individual colleges during the last five or six decades in the matter of concentration of the body of graduates into a few years. We may in a measure take the degree of this concentration as an indication of the homogeneity of the student body, and of the organization of the educational machinery that prepares the students for college. It will be noted that while there is the greatest difference in the degree to which the condensation has gone on in different colleges, there is, nevertheless, a distinct and uniform tendency towards this concentration, which must in every case be set down as a distinct advantage to the college. The ideal types may be said to be very nearly approximated by such curves as those of Yale, Plate VI., Adelbert and Dartmouth, Plate IV., and Alabama, Plate V. Such a curve as that of Dartmouth, which we may take as the type which all the other colleges more or less closely resemble, shows most clearly that the college has changed in sixty years from a place to which a young man might go for study at any age, to a place to which young men go as a matter of business, so to speak, and at a definite period of their life. In other words, the going to college has become a matter of social organization, with its very definite place in the life of the youth. The intermediate decades, which lack of space prevents our showing, present curves which show how gradually this change has come about. It seems, further, a safe conclusion to say that all the colleges that have not yet reached the high degree of concentration which some show are, nevertheless, distinctly destined to come to it, unless some unseen force changes their direction of development.

It should be noted, in passing, that an anomaly, such as the curve of Syracuse for 1850-59, is due to the small number of cases. There were but twenty-nine graduates in this decade.

Plate II. presents in graphic form the same facts that have been given in the tables. Division 'a' shows in the upper line, marked '1,' the average age of all graduates as presented in Table III., 'Average of Totals,' plus the data for decade 1900, so far as available, also referred to above. The second line, marked '2,' gives the actual median age of all graduates considered as students of one college. It will be noted that, while the median has remained practically uniform throughout, the average has varied, but with no marked tendency either up or down.

Plate II. 'b' presents the same facts as 'a,' except the units of comparison are now colleges instead of individual students. While, as would be expected from the small number of cases, the fluctuations are greater than in the 'a' division, the same absence of pronounced trend in either direction is easily observable.

There is one tendency in American education which it seems we may accept as established beyond cavil, viz., that for the future, the public high school will take the place of the old academy, as the institution in which the average boy will receive his training antecedent to entering college. In the days of our grandfathers, the prospective college student received his preparation for college either under the private instruction of his pastor, or in one of the academies of the time. In either case, the body of college-going boys was a highly selected one—a class who had both the tradition of the scholarly life and, to no small extent, the taste and opportunities to follow this tradition. Then, even more than now, the college turned out men whose future work was to be the ministry, law or medicine.

With the advent of the public high school and the growing tendency of colleges to accept its graduates for entrance to college courses, we should expect to find two or three changes in particular becoming manifest: First, we should expect to find the college-going students less selected along the lines of intellectual aptitudes and scholarly traditions; secondly, we should expect a greater scope of life employment among the college graduates; and thirdly, we should anticipate a natural advance in the age at which boys would go to college as a result of the above-named circumstances, with all that they imply. Now, our public school system is, for the most part, so constructed that the normal age for a boy to finish his high school course is in his nineteenth year, making his age of graduation from college between 22 years and 22 years, 11 months, inclusive.

From this point of view, it becomes important to examine our data with a view to finding out in how far these influences which would be expected to raise the age of graduation from college have been active over other conditions which have negatived them, or vice versa. Plate III. shows the percentage of students that actually graduated in all colleges under the age of 23 years, since 1850—the date at which the data for all our colleges become available. Comment is hardly necessary here. With the exception of decade 1860-69, which evidently shows the effects of the civil war, the trend has been unmistakably upwards. Even if we throw out the figures for 1900—which represent, as explained above, all the available data from the colleges that in 1890-99 furnished over 81 per cent, of all graduates—the trend is still unmistakably upwards.

Concerning the influences that have been instrumental in causing the marked rise in the median or average age of graduation in certain colleges in our list, it is not possible to speak with certainty for all. In the case of one or two, such as New York University and Bowdoin College, it would seem that the rise is due to an increase in the requirements for admission. In the case of certain other, pronouncedly denominational institutions, as DePauw and Syracuse, there is one element separable from perhaps others that may be surmised, which has played an important rôle. This is found in the decidedly high average or median age of those young men who go into the ministry. The following shows the conditions in the two institutions just named:

DePauw University (1).
Median of non-
Syracuse University (2).
Median of
Per cent, of
(1) (2) (1) (2) (1) (2)
1850-59 22 1 23 8 25 5 25 6 27.2 27.6
1860-69 23 1 23 3 23 3 24 6 22.8 41.6
1870-79 22 7 23 11 25 6 25 9 25.2 28.5
1880-89 22 11 23 3 25 3 25 6 25.4 31.7
1890-99 22 9 23 2 26 9 26 7 22.2 30.7

It thus appears that our medians for these two colleges as shown in Table II. would, with this element of disturbance removed, give quite different results. Thus the median of the last decade for DePauw would be lowered by Just twelve months; while that of Syracuse for the same decade, instead of remaining the same as that of fifty years before, would be lowered by nine months.

While I have not been able to work over the data for the other denominational colleges completely enough to give the results here, there are nevertheless many indications that a similar state of affairs prevails, though probably in different degree.

In conclusion, we may sum up our findings as follows: The increase in age of graduation from college in general has been tremendously exaggerated. It exists only for certain institutions; while others show a corresponding decrease.

The normal age of graduation, as our school system is constituted, is below twenty-three years and above twenty-two; our results show that more students graduate now within those limits than ever before; that the gradually organizing secondary education tends to make this percentage increasingly larger. (Nearly 85 per cent, of all graduates of the Johns Hopkins University in the twenty years since its founding to 1899 have been within these limits.)

If entrance into professional life is later than formerly, the cause must be sought elsewhere than in the college and preparatory school.

Whereas it was once possible for a boy to graduate from college at sixteen or even younger, though very few really did so, this is true no longer. But the young man now, as a consequence, leaves college with very much higher academic attainments, and but little if any older than was his father, or even his grandfather.

All colleges show, in different degrees, an increasing diminution of range in age of graduation. This shows that the secondary education is becoming better organized.

If now, the age of graduation which we have shown to be the prevailing one, viz., 22.5 years, be deemed still too old, three means of reducing this would seem to be possible: First, cut off one year from the college course, without lowering the entrance requirements; secondly, in view of the far greater efficiency of the secondary school, reduce the entrance requirements to college and, retaining the four! years' course, permit the boy to enter college, say, a year younger; thirdly, drop one year from the college course, increase the length of the actual weeks of residence and instruction to thirty-eight or forty, and endeavor to disabuse the mind of the average collegian of the belief that college is a place to dawdle and loaf four years for the sake of a degree that he does not earn, but which he generally gets Just the same. The college would then have a serious opportunity to prove its right to existence, and if it succeeded, the present dilettantism of college life would tend to disappear.

One further suggestion we may venture to make. Every boy that has the native capacity to do college work should be put into the high school in the fall after he is fourteen years old, regardless of whether he has done all the prescribed grammar school work or not. If he can not then get ready for college by eighteen, don't let him go to college. He is not cut out for the strenuous intellectual life.

  1. In Table I., decade '1770-79,' equals Dartmouth 1771-79; decade '1800-09,' equals Middlebury 1803-09; decade '1830-39,' equals Alabama 1832-39, New York University 1833-39, Oberlin 1837-39, Wesleyan 1833-39; decade '1850-59,' equals in Syracuse 1852-59. In each case the corrected year marks the date of the first graduating class. In decade '1890-99' Adelbert includes only the years 1890-95; New York University, 1890-94; Syracuse, 1890-98. In Alabama University there were no graduates for the years 1866-71 inclusive. During several of these years the university was closed. The data for the decade '1900- ' are as follows: Dartmouth, Oberlin, DePauw, each, class of 1900 only; Wesleyan, Alabama and Vermont, classes of 1900-01; Bowdoin. 1900-02. The whole number of cases in this 'decade' is 572. In reference to the degrees included in the investigation, I have attempted to use only A.B., Ph.B. and B.S. In a few instances the last named degree seems to be used as a semi-professional degree, implying for instance, that the student has taken an engineering, or some such course not purely 'cultural.' It seemed impossible to shut out entirely cases of the semi-professional degrees. The number of them is, however, too small to materially influence the results. In Dartmouth College the graduates of the Chandeler Scientific School are not included in the calculations, for the reasons above given. The justice of the exclusions above referred to is evident at once; for the examination is an attempt to show the changes that have come about in the college course as formerly understood. That is, when it did not include the study of a profession within itself, as several of the present courses do. Only young men have been considered in my inquiry. It is interesting, however, to note that if young women had been included in the investigation, the averages and medians would have, in almost every case, been materially reduced. In other words, the young woman is either more highly selected as a student, or she meets with fewer hindrances external to her work while going through high school and college. At any rate, whatever the cause or causes may be, the young woman graduates are, as a rule, younger than the young men in the same college. This subject is worthy of a separate inquiry.