TRENDS IN THE QUANTITIES OF EDUCATION: A PICTORIAL ESSAY Julian L. Simon, University of Maryland and Rebecca Boggs, Oxford University Abstract This essay shows in graphs that the amount of education youths receive has been increasing rapidly in the United States, and in the rest of the world. This represents an increase in individuals' capacity to produce economically and have a higher standard of living. The total amount of education has also been increasing because of both the increase per person and the larger populations. This implies an increase in the amount of knowledge that is created, which increases the standard of living for the community as a whole by increasing the productivity of individuals with given amounts of education. Perhaps the most impressive, exciting, and heartening trend - starting in the nineteenth century in the advanced countries, and continuing and speeding in the twentieth century in all the world - is the increase in the amounts of education youths re- ceive. This trend is crucial for broad economic development, of course. But in human terms it represents the opportunity for fulfillment rather than frustration for the many talented people of the world, and benefit to all humanity now and forever from the contributions to knowledge that these talented people will make. Despite the solid good news about education in almost all respects, however, the press and the educational establishment give the impression that the news is bad. Consider, for example, how a story on high school dropouts - considered important enough to be on the front page of The Washington Post - suggests that bad things are happening. The headline was "D.C. Dropout Rate Among Worst in U.S.", and a university president was quoted as saying "It is probably the most dangerous national security risk we're facing today". The president of the Carnegie Foundation for the Advancement of Teaching "explained" what he called the "high dropout rate nationally". His two reasons were "anonymity and irrelevance". We will return to the special topic of dropouts with data in hand, and you will then be able to evaluate for yourself the relationship of the news to the facts. The graphs in this paper document the trends, which are mostly so obvious that they need little commentary.1 TRENDS IN THE UNITED STATES Figure 1 shows the long-run trend in the proportion of children of school age enrolled in school from 1850 to the present. (The proportion among non-whites - mostly blacks - also is shown here because it is so eye-catching.) Figure 1 The data in Figure 1 understate the growth in education because the proportion of enrolled students who actually attended school on any given day was much lower in past years than now; that increase in attendance is shown in Figure 2. The length of the school term also has increased, also seen in Figure 2. The composite result of those two elements is the number of days in school per enrolled student. Figure 2 also combines the data for enrollment and for number of days attending school into a display that shows the number of days in school per year per person of school age. This is the best overall measure of the amount of education received up to the end of secondary school. Figure 2 Figure 3 shows the trends in the teacher-enrolled pupil ratio, and also shows that ratio adjusted for the number of students actually in school each day. If these data were combined with the composite data in Figure 2, we would have a measure of the amount of teacher time per child of school age per day or year. Figure 3 Data on attendance at and completion of secondary school are shown in Figure 4 - enrollment among youth aged 14-17, proportion of 17 year olds who graduate, and the proportion of the 25-29 year old cohort who have not finished high school. Figures 5 and 6 provide data on tertiary schooling. Figures 4-6 The summary statistic - median years of schooling achieved by those aged 25-29 - is displayed in Figure 7, separately for whites and blacks (combined data not available). And Figure 8 shows percentages with five years of school or less, and 4 years of high school or more. The data on amount of education measured in years undoubtedly are misleading - biased toward showing convergence - because there are important differences in the quality of the educations received by students of different races. But it is equally true that the apparent trends toward convergence are not entirely misleading, as shown by the earnings of blacks and whites of equal amounts of education measured in years. Figures 7 and 8 Before moving from the U.S. to the world situation, let us return to investigate how the newspaper story mentioned in the introduction could paint a bleak negative picture of the "dropout rate" when the trends in Figure 4 (also in Figures 5 and 6) are positive. Indeed, the diagram accompanying the newspaper story shows a positive trend toward less dropping out from 1972 to 1991. One cause of a negative impression is focusing on the Dis- trict of Columbia, which one may argue is a legitimate local negative story. But the story offers a negative outlook for the nation, too, as a whole; this is accomplished by quotes from supposed authorities. The factual positive trend is noted only in the last two paragraphs of the news story, for "balance". The facts presented would have been even more positive if the news- paper chart had been carried backwards even just a bit before 1972, as seen in Figure 4. TRENDS IN THE WORLD Long-run series for the poorer countries of the world are hard to come by. Figures 9, 10, and 11 provide aggregate attend- ance data for the world, and for the developed and developing parts of the world, for persons 6-11, 12-17, and 18-23 respec- tively. Figure 12 shows the skimpy, rough data on illiteracy rates by birth cohort in the developing countries. Figure 13 provides some long-run enrollment rates for selected countries. Figures 9-13 COMPARATIVE U.S. EDUCATIONAL EXPENDITURES AND ATTAINMENTS It is certainly difficult, and perhaps impossible, to find meaningful indices of how well a country is educating its people. Therefore, it is customary to examine crude measures of output such as years of education, and measures of input such as amounts of expenditures. Recognizing the crudity of such measures, let us briefly study a case in which the U. S. is said to be doing badly. Figures 14 and 15 show that in absolute money terms, the United States spends more per student at the primary and secondary levels (in purchasing power parity measures, which is a bit tricky) than does almost any other country; only Switzerland may spend substantially more, and only Sweden, Canada, and Denmark are similar (see Ram, 1995, for more data and analysis). This is exactly the opposite of the impression the story leaves with its focus on percentages rather than absolute amounts. Japan and Germany - which certainly have not done so badly economically - spend much less than the U.S. not only in dollar terms but in percentage terms. If one wants to connect economic success to this measure, the cases of Japan and Germany would hardly support the idea. Figures 14 and 15 At the tertiary level, the United States has a larger proportion of students attending than do most other countries. Combined with the expenditure per student shown in Figure 16, the United States clearly spends more per tertiary student, also, than does almost any other country. Figure 16 Even if it were true that the United States were at the bottom of the list in percentage terms--which is most certainly not so--this would not necessarily be bad evidence. As a person gets richer, s/he tends to spend a smaller proportion of income for certain important goods -- for example, food. On average, people in the richest countries--just like the richest families - -spend a smaller proportion of their income on necessities such as food than do poorer persons. Indeed, a small proportion of one's income spent on such things is an exceedingly good indicator of how well off one is in general. In aggregate, a poor country may spend 80% of its income for food, whereas a rich one spends only perhaps 25%. It would hardly be sensible to say that the rich country "lags" in food spending. And there is good reason to think that the same is true for education. A poor family certainly spends more for education as a proportion of its income than does a wealthy family, on average. Hence the data on proportion of income spent for education tell little about countries. More generally - and very importantly - money expenditures on the education of a given student are a very dubious indicator of the quality of the education that that student receives, if "quality" means some concept relevant to the preparedness of the student for adult life, economic and otherwise. One obvious defect is that teachers' salaries are a function of the income level of the society in which they teach; the "quality" of a teacher in the U. S. today, relative to that of a teacher in the U. S. one or two generations ago or of a teacher in Russia or South Korea or Israel is not appropriately measured by the relative salaries. There are also other very great defects with this measure. The absence of evidence of a connection within the United States and within Great Britain between educational expenditures per student and the results of the education - including data on the performances of students schooled in parochial schools and at home (see Beisner, 1990 for a brief review and references; see also Kiesling, 1994) casts even more doubt on the meaningfulness of this measure. It is discussed here only to dispell wrong impressions left by unsound analysis of the data themselves. If one is interested in how the United States compares against other countries, it would certainly be relevant to consider the total amount of education Americans have. The OECD book from which the above data come also includes a table of the educational levels of people 25 - 64 in various countries. Though this does not refer to current educational efforts, it does put the matter in perspective. Given that most Americans finish high school eventually, the best measure of total education is the people who have a college education. These data show that the United States is far ahead of any other. Figures 17-19 show how the U. S. has a much more educated adult population that does any other country - suggesting that the U. S. does not "lag". Figures 17-19SUMMARY AND CONCLUSIONS The amount of education that youths receive has been in- creasing rapidly in the United States, and in the rest of the world, over the periods for which we have data. This indicates an increase in people's capacity to produce economically, and therefore to have a higher standard of living for themselves. The total amount of education has also been increasing, both because of the increase per person and also because of the larger populations. This implies an increase in the amount of knowledge that is created, which increases the standard of living for the community as a whole by increasing the productivity of individu- als with given amounts of education. There is no obvious reason why these trends should not continue indefinitely. page 1 \article5 educart3 February 16, 1996 REFERENCES Beisner, Calvin, "Here's the One Thing Public Schools Fear Most", World, March 17, 1990, p. 12 (no volume number). Ferriss, Abbott, Indicators of Trends in American Education (New York: Russell Sage Foundation, 1969) Kiesling, Herbert J., "Reading the Report Cards: What Do 'State of Achievement' Reports Tell Us About American Education? - Review Essay", Economics of Education Review, Vol. 13, No.2, 1994, pp. 179-193. Ram, Rati, "Public Educational Expendutures in the United States: An Analytical Comparison with Other Industrialized Countries", Economics of Education Review, Vol 14, No. 1, pp. 53- 61. U. S. Department of Education, Digest of Education Statistics, NCES 92-097 (Washington: GPO, 1995) page 2 \article5 educart3 February 16, 1996 ACKNOWLEDGMENTS The assistance of Abbott Ferriss, Michael McEvoy, John Phelps, Ilya Somin, and George Washnis, is much appreciated. Guenter Weinrauch did his usual excellent work as research assistant, and prepared many of the graphs. Calvin Beisner gave us several valuable suggestions, and als prepared some of the graphs. Many of the U. S. graphs build upon Abbott Ferriss's excellent 1969 work. ENDNOTE 1. Concerning the sources stated below the graphs: After the original figures had been prepared, there appeared the excellent publication of the U. S. Department of Education (DOE), Digest of Education Statistics (1992 and annually thereafter), which contains several graphs similar to those shown here. For those where there was substantial similarity to the eye and therefore leave the same impression, the graphs here were not redrawn. But the sources given for the DOE figures are given below the figures here. The reason is that the figures drawn here were often cobbled together by a variety of hands from a variety of sources (in many cases beginning with the volume of Ferriss), which are not easily accessible, and which in the case of Ferriss already is a secondary source. The DOE volume could have been given as a sole source. But the person whose needs we have in mind is the person who may wish to extend these figures at some later time, or otherwise modify them, and the most useful sources to such a person are those that will be authoritative, accessible, and especially that later volumes of them will contain the newer data. Hence the references given, taken in many cases from the DOE volume, have been chosen with this object in mind. page 3 \article5 educart3 February 16, 1996 NOTE FOR FIGURE JLS.12.ls1 Note: The two lines represent estimates made from two surveys taken in different years, 1970 and 1990. Both provided estimates for the groups born from 1945 to 1953 page 4 \article5 educart3 February 16, 1996 NOTE FOR FIGURE JLS19C Note: The countries with zero entries are so listed in the original. page 5 \article5 educart3 February 16, 1996 NOTE FOR FIGURES JLS16, 17,and 18 Note: Data are not available where the entries are blank. NOTE FOR FIGURES JLS16, 17,and 18 Note: Data are not available where the entries are blank. NOTE FOR FIGURES JLS16, 17,and 18 Note: Data are not available where the entries are blank. page 6 \article5 educart3 February 16, 1996