Click for next page ( 29


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 28
2 Characteristics of Doctorate Recipients In the first chapter, we were concerned with the numbers of PhD's, as they varied over time, by field, and by sex. We turn now to the charac- teristics of the doctorate recipients themselves— those characteristics that can be tabulated from the data of the DRF. These appear to be of pri- mary importance regarding the education and employment of these people—particularly to the educational institutions and to the agencies that provide support for graduate education. These characteristics, in the order in which they will be described, concern: 1. The educational background of the families from which they come. 2. Citizenship and racial/ethnic identifica- tion. 3. Age and the time lapse between baccalau- reate and doctorate degrees. 4. Master's degrees. 5. Field switching between the baccalaureate and doctorate levels. 6. Geographic migration, region by region within the United States, from high school to PhD. HIGHLIGHTS • Educational Background. The general popu- lation of the United States has become steadily better educated over the past century, at the rate of a little less than two grade levels per generation. The PhD's have come from families at the leading edge of this educational wave— from families that were, on the average, one generation ahead of the general population. There are significant sex differences: The women PhD's come from slightly better-educated families than do the male PhD's. Field differ- ences also exist but are decreasing in magnitude. The pattern of all of these changes makes a fas- cinating mosaic. • Citizenship. One in seven PhD's awarded in the United States is to a non-U.S. citizen. The proportion varies profoundly by field: for- eign citizenship is highest in male-dominated agricultural sciences (33 percent), engineering (28 percent), and medical sciences (21.5 per- cent) , and lowest in education (5.4 percent) and psychology (5.2 percent), in which the proportion of women is much higher. Thus the field differ- ences can be said to explain a large part of the overall sex differences: 15 percent of the male PhD's and 10 percent of the female PhD's are non- U.S. citizens. • Racial/Ethnic Identification. Data on racial/ethnic composition of the doctorate recip- ients has only recently become available. It varies by field, and hence, to some extent, by sex. Overall, including U.S. and foreign citizens but omitting those for whom racial/ethnic data are unavailable, 87.7 percent of recent PhD's are white, 3.4 percent are black, 0.5 percent are American Indians, 1 percent are Spanish Americans, Mexican Americans, or Chicanos, 0.2 percent are Puerto Ricans, and 7.2 percent are Orientals. Blacks and American Indians tend to be concen- trated in education, and Orientals in the EMP fields. • Age. The typical PhD is about 30 years old at graduation—younger in the sciences, older in the nonsciences, particularly education. Age at baccalaureate and age at doctorate tend to show the same pattern of field differences, but there

OCR for page 28
29 is less spread at the BA level. Age at PhD is therefore determined principally by time lapse between the baccalaureate and doctorate. BA-PhD time lapse has increased over the past half- century, but the major fluctuations were those induced by World War II and its interruption of the educational progress of both men and women, but particularly the men. • Waster's Degrees. Except in chemistry, most PhD's also have master's degrees. In chem- istry, 41 percent have the degree; in physics, 64 percent; in the biosciences and the medical sciences, 65 percent; in psychology, 77 percent; in the earth sciences, 78 percent; in mathematics, 79 percent; in the social sciences, 83 percent; in the professions, 86 percent; in the humani- ties, 87 percent; in engineering, 89 percent; in the agricultural sciences, 90 percent; and in education, 97 percent. The percentages are typically higher for women than for men, the ex- ceptions being the earth sciences, engineering, and agricultural sciences. • Fields at BA and PhD. Field switching, for the doctorate-bound population, results principally in flows from mathematics, physics, chemistry, engineering, the agricultural sciences, and the humanities to the biosciences, the earth sciences, and education. The other fields are in relatively close balance overall, but for the women there is a particularly strong movement out of the professions1 and the medical sciences.2 In this report each field is con- sidered in terms of its donor/receptor character- istics: the extent to which it "donates" its baccalaureate recipients to the various "receptor" fields at the doctoral level. • Interregional Migration. Most PhD's earn their doctorates in the same geographic regions in which they graduated from high school and from college. The regional shifts have varied over time and are a function of the relative strength of each region at the secondary, higher education, and graduate levels and population. Patterns of migration are explored in terms of "donor" and "receptor" regions, at the HS-PhD level and BA-PhD level. SOCIOECONOMIC BACKGROUNDS OF DOCTORATE RECIPIENTS Potentially, there are a number of indicators that could be used to describe the socioeconomic backgrounds of doctorate recipients. However, as a practical matter, the only indicator avail- able in the DRF is the level of education attained by the parents of the PhD's. Fortunately, this is an important indicator for this particular group, distinguished as it is from the general population primarily by its educational attainment. 'The professions include business administration, home eco- nomics, journalism, theology, law, social work, library science, and the speech and hearing sciences. The medical sciences include medicine and surgery, dentistry, veterinary medicine, hospital administration, parasitology, pathology, pharmacy, and pharmacology. It is of course to be expected that PhD's come mostly from the better-educated families. The extent of the difference in the educational spectrum from which PhD's come, as compared with the general population, was explored in Profiles of PhD's in the Sciences, published by the NAS in 1965. That study compared the educational levels of the general population with those of the parents of the PhD's who graduated over the period from 1935 to 1960. Because PhD's are, on the average, about 30 years old at the time they take the doctorate, and because their par- ents are, on the average, assumed to be about 30 years older than that, the time differential between the birth of the parents and the year in which the PhD's graduate is assumed to be 60 years. It is this time differential that was used to compare the PhD's and the general popula- tion in the 1965 study. UPDATE AND NORMATIVE FRAMEWORK It is now possible to update and extend the earlier study. A sample of 10,000 PhD's was used in the 1965 study, drawn from the graduation co- horts of 1935, 1940, 1945, 1950, 1955, and 1960. At the present time, complete data are available for the more recent graduates, here divided into four cohorts, the PhD's of 1963-1965, 1966-1968, 1969-1971, and 1972-1974. Census data from the decennial censuses for 1940-1970 provide informa- tion on the educational levels attained by the general population, typically divided into 10- year age cohorts. Educational level is recorded at nine steps of attainment: no formal education; grades 1-4; grades 5-7; grade 8; 1-3 years of high school; high school graduation; 1-3 years of college; college graduation; and postcollege training. In the tables and graphs to followr some discontinuities, showing up as jagged lines in percentile graphs of educational attainment, will be found. This is in part a result of the particular steps of attainment that were employed, but it is also due to the fact that, historically, generally accepted termination points of formal education have been eighth grade, high school graduation, and college graduation. In the case of the PhD's in the DRF, a slightly different set of educational attainment points was used (third grade instead of fourth; sixth grade instead of seventh; and an additional level at the top, differentiating master's degrees and the doctorate). However, the data sets are compatible, and meaningful comparisons are provided, using the assumption described above to define the birth cohorts of the parents of PhD's. In examining the graphs, particularly Figures 27 and 28, a slight truncation of the norm for the general population will be noted for the most recent cohort. This is because data were available in 1970 for persons age 25 and up, but some of them (more men than women) had not com- pleted their formal education at that time. The limitation is slight and does not interfere with the usefulness of the data, except for post- baccalaureate degrees.

OCR for page 28
30 TABLE 9 EDUCATIONAL ATTAINMENT OF THE UNITED STATES POPULATION, BY BIRTH COHORT AND SEX (Averaged Data from Censuses of 1940,1950, 1960, and 1970)* Educational Level Attained Year of Birth Males No education % Before 1866 1866-1875 1876-1885 1886-1895 1896-1905 1906-1915 1916-1925 1926-1935 1936-1940 10.13 10.13 8.28 8.28 6.73 6.73 5.40 5.40 2.79 2.79 1.46 1.46 1.02 1.02 0.95 0.95 0.91 0.91 Grades 1-4 t 18.63 28.76 18.48 26.76 16.44 23.37 14.29 19.69 9.75 6.33 3.85 4.87 2.49 3.44 1.52 2.43 C» 12.53 7.80 Grades 5-7 % 23.03 51.79 22.84 49.60 22.18 45.54 21.73 41.42 19.92 32.45 15.41 23.20 9.90 14.77 7.30 4.93 7.36 Ct 10.74 8th grade t Ct 33.00 84.79 30.18 79.78 29.69 75.23 28.11 69.53 26.48 58.92 19.74 42.95 12.60 27.36 8.79 19.52 5.95 13.31 High school, t 1-3 years C\ 4.80 89.59 6.72 86.49 8.91 11.53 15.94 74.86 20.25 63.20 21.19 48.55 19.74 39.26 17.90 31.12 84.13 81.06 High school t graduate C% 4.99 94.58 6.55 8.06 92.19 9.32 12.54 87.40 20.26 83.46 29.37 77.92 31.77 71.03 37.13 68.24 93.04 90.36 College, % 1-3 years Ct 2.53 3.22 3.62 4.74 95.12 6.08 93.47 7.96 10.05 87.97 11.50 82.53 13.44 81.69 97.11 96.26 95.81 91.42 College % graduate c% 2.01 99.12 2.52 2.88 98.69 3.14 98.26 3.84 97.32 4.31 95.73 6.20 94.16 8.67 91.20 8.64 90.33 98.78 Graduate/pro fes- t sional school Ct .87 1.22 100.00 1.31 100.00 1.73 100.00 2.69 4.26 5.83 8.80 100.00 9.67 100.00 100.00 100.00 100.00 100.00 Median 7.27 7.51 7.65 7.81 8.16 9.54 11.55 11.84 12.01 Mean 6.39 6.77 7.17 7.65 8.59 9.67 10.76 11.53 12.01 Females No education % Ct 9.44 9.44 6.91 6.91 5.87 5.87 5.19 5.19 2.62 2.62 1.23 1.23 0.82 0.82 0.86 0.86 0.81 0.81 Grades 1-4 t Ct 14.88 24.32 14.18 21.08 13.02 18.89 11.13 16.32 7.80 4.56 5.79 2.65 3.47 1.70 2.56 1.18 1.99 10.42 Grades 5-7 t ct 21.54 45.86 21.52 42.60 20.83 39.72 20.27 36.59 18.32 28.73 13.90 19.69 8.59 12.06 5.65 8.21 4.04 6.03 8th grade V 35.26 81.12 32.24 74.84 29.99 69.70 27.76 64.35 24.68 53.41 18.53 38.22 11.66 23.72 7.19 5.12 11.15 Ct 15.40 High school, % 1-3 years C% 6.34 87.46 8.78 83.62 11.43 81.13 13.75 78.10 17.11 75.52 21.13 21.93 45.65 21.90 37.30 20.22 31.37 59.35 High school t graduate C% 8.01 95.47 10.25 93.87 11.53 92.65 12.50 90.60 16.77 87.29 24.91 84.26 37.62 83.27 42.41 79.71 45.26 76.63 College, % 1-3 years C% 2.81 98.29 3.82 97.69 4.53 97.19 5.89 96.49 7.73 95.02 9.19 93.45 10.06 93.33 11.38 12.86 89.49 91.09 College % graduate Ct 1.50 99.79 1.98 99.66 2.32 2.62 99.11 3.56 4.03 97.48 4.42 97.74 5.89 96.99 7.29 99.50 98.58 96.78 Graduate/profes- * sional school Ct 0.21 0.36 100.00 0.51 100.00 0.89 100.00 1.42 100.00 2.52 2.27 100.00 3.02 100.00 3.22 100.00 100.00 100.00 Median 7.62 7.73 7.84 7.98 8.36 10.17 11.62 11.80 11.91 Mean 6.75 7.25 7.60 7.99 8.89 9.91 10.73 11.30 11.67 See text for censuses contributing to each average. ^C» • cumulative percent. The general population educational attainment data are shown in Table 9. The percentage com- pleting each level, and the cumulative percentage up to that level, is shown for each birth cohort, for men and for women. Additionally, means and medians, by cohort and sex, are given. The data on mean educational levels from this table are plotted in Figure 24, which also shows comparable data for the educational levels of the parents of PhD's, for the birth cohorts for which data are available. In the case of both the general population (shown as heavy lines) and the PhD population (shown as lighter lines), the data for males are given in solid lines and the data for females in dashed lines. It is apparent ^For those who may wish to compare the data of Table 9 with other sources, it should be noted that the columns of this table usually combine data from two or more censuses to obtain more stable percentages. This is particularly important at the extremes of the distributions, where data are sparse. The census data available were from rather small samples, rather than complete figures. The pre-1866 data were taken solely from the 1940 census; 1866-1875 and 1876-1885 data from the 1940 and 1950 censuses; 1886-1895 data from the 1940, 1950, and 1960 censuses; 1896-1905 and 1906-1915 data from the censuses of 1950, 1960, and 1970; 1916-1925 data from the censuses of 1960 and 1970; and the rest from the 1970 census alone.

OCR for page 28
31 13 r- Females, Age 25 and Up, U.S. General Population Males. Age 25 and Up, U.S. General Population 6 - The parents of PhD's have been about 30 years ahead of the general population, on the average, in educational level (average grade attainedl I I 1860 1870 1880 1890 1900 1910 1920 1930 1938 ASSUMED MEAN YEAR OF BIRTH FOR COHORT SOURCE: NRC, Commission on Human Resources FIGURE 24 Educational level of parents of U.S. native PhD's compared to U.S. general population, by year of birth. that over the 75 years shown here, there has been a steady progression of educational attainment. The trend for the two sexes is similar, but prior to 1920 the mean for women was higher than that for men, whereas the reverse is true for the more recent cohorts. PARENTS AND POPULATION NORMS The educational level- of the parents of PhD's is in marked contrast to that of the general popu- lation, as far as the means in Figure 24 are concerned. From the earliest cohort shown until the beginning of the twentieth century, the par- ents of native-born U.S. PhD's averaged just under high school graduation as their highest level of educational attainment."4 Meanwhile, ''Parents of U.S. natives only are included here, both because of the difficulty in equating educational levels across cul- tural lines and because of field and cohort differences in percentage of persons of foreign origins. Had they been in- cluded, some marked distortions would have been produced. the general population norm moved up from about the seventh grade to about the eighth grade. From the beginning of the present century, the average of parents of PhD's moved up approxi- mately parallel to the change in the general population norm. It is interesting to note that, prior to 1900, the mean educational level of the mothers of PhD's was below that of the fathers, but in the more recent cohorts the dif- ference in means has vanished. The difference in distribution of educational attainments has not vanished, as will be seen, thus illustrating a limitation of mean values to describe a popu- lation characteristic.

OCR for page 28
32 13 12 11 !» —i ui ui o tr 6 The U.S. mean educational level has increased steadily; the median jumped quickly after 1900, as "typical school leaving" responded to compulsory education laws I I I I 1860 1870 1880 1890 1900 1910 1920 1930 1938 YEAR OF BIRTH SOURCE: U.S. Decennial Censuses of 1940, 1950. 1960, and 1970 FIGURE 25 Changing educational level of U.S. population: means versus medians. A CHANGING EDUCATIONAL SPECTRUM The difference between means and medians may be noted in examining the data of Table 9. It is illustrated graphically in Figure 25 for the general population. Here we see again the pro- gression of means over the same period as shown in Figure 24. Median data are also shown and, by contrast to the means, show sharp changes during the first 20 years of the present century. The medians rise at a very modest rate until the beginning of the twentieth century, when they shoot up rapidly, then rise slowly after 1920. This is an effect due to the quite rapid change of the middle section of the population—a move

OCR for page 28
33 1876-1885 40 30 20 10 0 1,: LJ I 1 40 30 20 2 10 I. I I 40 30 20 10 0 1916 1925 o 1- 40 0 30 £ "• 20 10 0 40 30 20 10 0 1936-1940 40 30 20 10 0 0368 12 16 20 GRADE LEVEL 0368 12 16 20 GRADE LEVEL SOURCE NRC, Commission on Human RtsourcM, based on BurMu of Census data FIGURE 26 Distributions of educational attainment of general population age 25 and up, by birth cohort and sex. from a norm of eighth grade graduation to a norm of high school graduation. The median is af- fected by changes around the midpoint only, where- as the mean is affected by changes at any point in the educational scale. Figure 26 shows fre- quency diagrams of the percentage of the popula- tion, by sex, at each educational level recorded in the census statistics, for selected birth co- horts, from those born between 1876 and 1885 to those born between 1936 and 1940. The peaks of the distributions shift, in the first 2 de- cades of the twentieth century, from eighth grade to twelfth grade. In the 1936-1940 co- hort the grade level range has been extended by incorporation of data calculated from DRF to supplement the census data.

OCR for page 28
34 100 g 1 o o §- 5 o o 2 E O (9 60 O 40 I co 20 No Formal Education The "grade level mix" for the U.S. adult male pop- ulation has shifted strongly; tha norm now is high school graduation or beyond 1-3 Years of High School Some Postcollege Training J I I 1860 1880 1920 1900 BIRTH COHORT SOURCE: NRC, Commission on Human Resource, bawd on Bureau of Census data FIGURE 27 Changing educational spectrum of U.S. male population. 1938 GROWTH CURVES OF EDUCATIONAL ATTAINMENT A sex difference is visible in the frequency polygons of Figure 26 chiefly by way of a larger proportion of men who have gone to college. The changes over time in educational attainment are not as easy to see in Figure 26 as in the next graphs, which show time changes in the various levels of educational attainment. The proportion of the population which has had no formal educa- tion decreases, for both men and women, from about 10 percent to about 1 percent in Figures 27 and 28, which are taken from the data of Table 9. The proportion who are high school grad- uates, but who go no farther than high school, is shown as the shaded area in the center of the graph. For the men, this area increases gradually and rather regularly; for the women there is an al- most explosive growth after the beginning of the twentieth century. The shaded area near the bot- tom of the graph in both pictures indicates those who have completed baccalaureate degrees but no more. This is somewhat larger for men than for women, but it is the portion beyond the baccalau- reate that shows the greatest sex difference. In the most recent cohort (where data were incom-

OCR for page 28
35 100 High school graduation-hut no more-has tended to become the norm for U.S. women 1860 1880 1920 1900 BIRTH COHORT SOURCE: NRC, Commission on Human Resources, based on Bureau of Census data FIGURE 28 Changing educational spectrum of U.S. female population. 1938 plete in the 1970 census), the proportion of men is almost 10 percent; for women it is only slightly over 3 percent. The curves for all educational levels progress rather smoothly, with the exception of the very rapid shift in high school graduations after 1900. This is probably the effect of changes in the compulsory education laws. These state laws, enacted mostly during the last half of the nineteenth century, began to have a marked effect at the high school level in the beginning of the twen- tieth century. At that point most state laws required attendance only up until age fourteen; by 1920, age sixteen was a more typical school- leaving minimum. Because these state laws were not all enacted simultaneously, and because of inevitable lags in enforcement, the effects were not sudden—although as noted earlier, the ex- pansion of the women, high-school-graduate-only group is quite rapid, because a much smaller proportion of women than of men go on to college. sSee A. W. steinhilber and C. J. Sokolosky, state Law on Compulsory Attendance, Publication OE 23044, circular 793 (Washington, D.C.: USOE, 1966). (Superintendent of Documents Catalog FS 5.223:23044.)

OCR for page 28
36 0.5 1 18 PERCENT ABOVE INDICATED GRADE LEVEL 10 30 50 70 90 95 98 99 99.5 16 16 12 11 Q < K T T 1 I I i r T T T I I I College Graduation The spectrum of educational levels has shifted steadily for U.S. males, over periods for which census data are available 0 99.5 99 98 95 90 70 50 30 10 5 2 1 0.5 PERCENT BELOW INDICATED GRADE LEVEL (Note Normal Probability Spacing) SOURCE: NRC, Commission on Human Resources, based on Bureau of Census data FIGURE 29 Percentile graphs of educational levels attained by U.S. males age 25 and over, by decade of birth. PERCENTILE NORMS OF EDUCATIONAL ATTAINMENT Up to this point, we have considered means/ medians, frequency distributions, and growth curves of the educational levels of the general population. In order to put the data into a form that will facilitate comparison with the educational spectrum of the fathers of PhD's, Figure 27 has been recast into percentile terms, with one percentile curve for each birth cohort, in Figure 29. A similar set of curves could be drawn for the general population of women, as a normative frame for the mothers of PhD's. In both cases, the progression of the birth cohorts is seen as a march of the curves across the page from left to right. The curves for women (not shown because they are so similar as to be redun- dant) vary only in that smaller percentages achieve the higher levels of education, although at the elementary education levels, the percent- age of women at each grade level is slightly higher than that for men. The percentile data are plotted with normal probability spacing, which provides for equal intervals in terms of standard deviation units. This compresses the percentages around the middle of the distribution and expands the percentages at the extremes. In spite of this midrange compression, the greatest

OCR for page 28
37 0.5 1 PERCENT ABOVE INDICATED GRADE LEVEL 5 10 30 50 70 90 95 98 99 99.5 ui > Q < DC O f I -V~~ t I College Graduation Fathers of PhD's of 1969-1971 (Assumed Birth 1906-1915) High School Graduation U.S. Male Population Born 1906-1915 Equivalent birth cohorts of the general male population and the fathers of I'hD's are close together at the lower educational levels, diverge at higher levels 99.5 99 98 95 90 70 50 30 10 5 2 1 0.5 PERCENT BELOW INDICATED GRADE LEVEL (Note Normal Probability Spacing) SOURCE: NRC, Commission on Human Resources FIGURE 30 Educational attainment of fathers of PhD's, by birth cohort, compared with the norms of Figure 29. changes are shown at about this point between the eighth and twelfth grades. Because it is the upper educational levels that are the primary concern with respect to the parents of PhD's, this method of normal probability spacing per- mits a clearer view of the changes where they are most relevant to the present study. If the percentiles had been plotted as equal intervals, the result would have been a tight compression at both extremes of the distribution, minimizing the most relevant data. Figure 29 provides a normative frame for interpreting the data on the educational attain- ment spectrum for the fathers of PhD's. This is done in Figure 30, where a heavy black line has been used to represent the general population curve for the birth cohort of 1906-1915, and a dashed line to represent the fathers of PhD's who were their contemporaries—the fathers of the PhD's of 1969-1971. A similar comparison could be made for the mothers of PhD's of the same era, compared to the general population of women, but the data are too nearly redundant to justify a separate graph. In both cases, in spite of minor sex differences, one may say as a rough generalization that the parents of PhD's are about one generation ahead of the gen- eral population in educational attainment.

OCR for page 28
38 0.5 1 PERCENT ABOVE INDICATED GRADE LEVEL 10 30 50 70 90 95 98 99 99.5 Ul Q oc o I I I I I I I Fathers of PhD's of 1969-1971 (Assumed Birth 1906-1915) High School Graduation U.S. Female Population - Born 1906-1915 Mothers of Ph D's of 1969-1971 (Assumed Birth 1906-1915) U.S. Male Population Born 1906-1915 3 - 99.5 99 98 95 90 70 50 30 10 5 2 1 0.5 PERCENT BELOW INDICATED GRADE LEVEL (Note Normal Probability Spacing) SOURCE: NRC, Commission on Human Resources FIGURE 31 Comparison of educational attainment spectra of males and females in general population and parents of PhD's. SEX DIFFERENCES IN EDUCATIONAL ATTAINMENT To summarize the comparison of the data on parents of PhD's as compared to the general population, and to present data for both males and females, Figure 31 shows four percentile curves. The heavy lines are those for the gen- eral population, the lighter lines for the par- ents of PhD's, and, in both cases, solid lines represent data for men, dashed lines data for women. For both the general population and for parents of PhD's, there is a crossing-over of the men and women's graphs at the high school level. The difference, however, is greater for the par- ents of PhD's than it is for the general popula- tion. In both comparisons, the curve for men is above that for women at the higher education level but below at the elementary school level. The data for the various grade levels for fathers of PhD's are given in Table 10 and for mothers in Table 11. In both tables, data are given sepa- rately for the female PhD's and the male PhD's and for both combined. At the bottom of the table, the summary statistics are provided: means, standard deviations, and the percentile points 10, 25, 50, 75, and 90. The sex differ- ences here provide an interesting study and will be examined in more detail in the graphs to follow. The interesting new information shown here is that the progression of the cohorts con- tinues, for both the mothers and the fathers of the PhD's, for the recent cohorts. The mean data shown here are shown graphically in Figure 24. The data of Tables 10 and 11 show that the same progression given for the means holds also for the other portions of the educational spectrum.

OCR for page 28
FIELD OF DOCTORATE 100 50 O _l UJ X u UJ U oz £ S cc i — u O 10 Math Phvxcs Chem Earth Engr Agr Med Bio Psv SocSci Hum Prof Educ Chem *** — Ph Hum Math Earth Agr Psy ~ Bio SS Educ - Med Prof Except in medical sciences, the professions, and education, most eventual PhD's stay in their baccalaureate fields for the doctorate. but the patterns are varied Bio Hum Hum - Chem Hum SS Engr Bio Chem Hum Engr — Phys — Engr Agr Prof Engr SS Psy Phys SS Prof Math Engr Chem Educ SS Educ Educ Math Bio Math Math Educ Bio Med Psy ~ Chem Agr Educ Hum Prof ^ Educ Prof Med Math Phys Hum Bio P.T Math Hum Hum SS eauc SS Math Chem Chem Chem Med P«V Engr Engr Math Agr Hum Bio Agr Engr f'°' Engr Chem -Math M*ri — SOURCE NRC. Commission on Human Resource! FIGURE 46 Relative frequency of various BA source fields, for each PhD field. nations in logarithmic form, to spread out those that would otherwise be too close together for legibility. Thus, in the case of mathematics, in column 1 of Figure 45, mathematics as a PhD destination appears near the top but still is only at the 56 percent point. Moving down the column, we note that about 15 percent of mathe- matics baccalaureates take doctorates in educa- tion, about 7 percent in the social sciences, 5 percent each in physics and engineering, 3 per- cent each in the humanities and biosciences, 2 percent each in psychology and the professions, and 1 percent each in the earth sciences and chemistry. In a similar manner, one may look down each succeeding column and note the percentage who remain in the field of undergraduate major and the percentage who switch to other fields. By reference to Table 24, a more exact statement of the percentages is available. Mathematics, as it turns out, is one of the lowest of the baccalaureate fields in retention of its gradu- ates through to the doctorate. Alternatively, it can be described as one of the best as a basis for getting a PhD in a variety of fields. High proportions of mathematics majors go into other fields, partly as a function of the trans- ferability of skills, and partly as a function of the relative size of the various fields. The contrasting sizes of the fields of earth sci-

OCR for page 28
68 ences and chemistry, both near the bottom of the mathematics column, brings the latter considera- tion into focus. Looking at the other fields of baccalaureate as contributors to their own fields at the PhD level, one sees that the earth sciences and education retain a high proportion (over 80 per- cent) through to the doctorate. Engineering (79 percent), the biosciences (75 percent), physics and psychology (73 percent each), and chemistry (69 percent) are intermediate, and the other fields are much lower in retention rate. It must be recognized that "retention rate" is a function of the breadth of the field and that in important ways fields designated here are not uniform in "breadth," although there is no way that one can define breadth objectively and quan- titatively. The transferability of skills learned in undergraduate training is an important factor. No doubt the ubiquitousness of the need for mathe- matical skills accounts in large part for the number of persons leaving mathematics as a spe- cialty and moving to other fields where their mathematics skills can be utilized. Another fac- tor in this particular case is the fact that mathematicians per se have little other than the academic area for employment, whereas by switching, they find more fields of application. The relatively high proportion of math BA's going into education undoubtedly represents a recognition that opportunities to teach mathe- matics and do research in mathematics are limited at the university level. If one majors in educa- tion, more opportunities open up in colleges, junior colleges, and even in high schools—perhaps for those with teaching skills and interests but less aptitude for research in mathematics. Physics as a baccalaureate source field con- tributes, not unexpectedly, to engineering about 8 percent of its graduates. Bioscience (4 per- cent) comes next, in large part, no doubt, because of the development of the growing field of biophysics. Mathematics and physics have a great deal of overlap in terms of skills learned and required, and mathematics absorbs almost 4 percent of physics majors. Other destination fields include the earth sciences and education (3 percent each) and the social sciences and humanities (1 percent each). Chemistry contributes a high proportion of its baccalaureates to the biosciences (18 per- cent) —a tribute to the size of the biochemistry field. Almost 4 percent of BA-1evel chemists go into education and 2 percent or fewer into engineering and medical sciences; 1 percent or less enter other fields. The earth sciences, as noted earlier, have the highest retention rate, but still contribute 4 percent of their graduates to education, 3 percent to the social sciences, 2 percent each to the biosciences and engineering, and 1 percent to the humanities. Engineering and physics, as noted earlier, have a reciprocal relationship, and physics is the major nonengineering destination field (5 per- cent) for engineering graduates; approximately 3 percent go into mathematics, chemistry, and the professions and between 1 percent and 2 per- cent into four other fields: earth sciences, biosciences, social sciences, and education. The agricultural sciences have an understand- ably close relation to the biosciences: 22 per- cent finish with bioscience PhD's. The social sciences get 6 percent, perhaps because of a cer- tain degree of ambiguity regarding the classifi- cation of agricultural economics. Education also claims 6 percent—undoubtedly primarily as teachers of agriculture. No other field takes over 2 percent. The medical sciences contribute 12 percent of their number to chemistry as a PhD destination field, probably concentrated mainly in Pharmaceuticals. Bioscience gets 23 percent; education, 13 percent; and psychology, the agri- cultural sciences, and the social sciences, 3 percent each. The net result is that only 41 per- cent of those with baccalaureates in the medical science fields take doctorates in this field. A certain degree of ambiguity attends this finding, however, since the coding of foreign pre-PhD degrees in this field involves some uncertainty and in the early 1960's MD degrees were coded here in a combined "baccalaureate and first pro- fessional" category. The biosciences have a high retention rate, but still about 8 percent go into education at the doctorate level, followed by 5 percent each to the agricultural sciences and the medical sci- ences, and 2 percent to psychology. Psychology, as might be expected, is closely related to education; about 18 percent of psychology majors end up with education doctorates. About 3 per- cent go into the social sciences, and 2 percent each into the humanities and the professions. The social sciences contribute about as many of their graduates to education (18 percent) as does psychology but a much higher proportion (7 percent) to the humanities, 6 percent to the professions, and 4 percent to psychology (an al- most even exchange). Of the humanities baccalaureates, over 20 percent finish in education, about 7 percent in the social sciences, 5 percent in the professions, and 4 percent in psychology. The "professions" are a very diverse set of fields, including theology, business administration, home economics, law, journalism, speech and hearing sciences, social work, and library science. The PhD field destinations are also diverse, including only 40 percent to the "professions," 25 percent to edu- cation, 17 percent to the social sciences, 10 percent to humanities, and 4 percent to psychol- ogy. Education, as noted earlier, has a high retention rate, but still 6 percent of education majors complete doctorates in the humanities, and about 3 percent each in the biosciences, the social sciences, and psychology. The psychology-education exchange is predominantly a one-way street. There is an additional row on the baccalau- reate side of Table 24 that is not shown on the chart of Figure 45. That row is for unknown baccalaureate fields. These range from about 2.5 percent to a little over 5 percent entering each PhD field, with the exception of the medi- cal sciences. As noted earlier, there is some ambiguity about the medical sciences at the "baccalaureate" level, and this is probably the

OCR for page 28
67 reason for the deviation of the medical sciences from all the others in the row for "baccalaure- ate field unknown." BACCALAUREATE SOURCE FIELDS As mentioned earlier, one may look at the field- switching phenomenon from an entirely different perspective: backwards from the doctorate fields to see what source fields contribute to each of the PhD disciplines. This is shown in diagram- matic fashion in Figure 46. Here it is immedi- ately apparent that each field is its own best supplier by a much higher margin than one would expect from the data of Figure 45. Mathematics supplies three out of four of its own PhD's, tak- ing 10 percent from engineering, 5 percent from physics, and 2 percent each from the humanities and education. The transferability of skills is undoubtedly a major factor in this pattern— fields other than engineering and physics are unlikely to require the development of mathemati- cal skills sufficient to permit their graduates to switch to mathematics as a doctorate-level discipline. A few make it, but undoubtedly because of special interests and choice of elec- tives, rather than by reason of required training. A similar and reciprocal set of relationships is found for the source fields for physics. Engineering contributes about 12 percent, mathe- matics about 4 percent, and chemistry and the humanities 1 percent each. Chemistry is even higher than mathematics and physics in the extent to which it draws on its own baccalaureate field for future doctorate recipients. It does, how- ever, draw also on engineering (4 percent), medical sciences (2 percent), biosciences (1 per- cent) , and the humanities (1 percent). The earth sciences, which had the highest retention rate from BA to PhD, is lower than any other natural science field as a source field for its own doctorates—no doubt because, as an undergraduate field, it is very small. It draws extensively from the other sciences, physics (9 percent), engineering (7 percent), math, the biosciences, and chemistry (about 4 percent each), and less on other fields (humanities, 2 percent; social sciences, 1 percent; agricultural sciences, 1 percent). Engineering is highly self-contained, but does draw about 4 percent of its doctorates from physics, 2 percent from mathematics, a little less than 2 percent from chemistry, and about 1 percent from the humanities. Agricultural sciences as a PhD field draws about three-fourths of its members from under- graduate majors in agricultural sciences, but it also draws heavily on the biosciences (11 per- cent) . Education and chemistry each contribute 3 percent, and the medical and social sciences about half of that. The ambiguities in the medical sciences as a first-level field do not apply at the doctorate. This field includes veterinary medicine, parasi- tology, pharmacology, pharmacy, pathology, envi- ronmental health, public health and epidemiology, hospital administration, and nursing, as well as "other" and "general." It is not surprising, therefore, that the source fields for the medi- cal sciences are diverse: 11 percent come from chemistry, 23 percent from the biosciences, 3 percent each from the agricultural sciences and the humanities, 2 percent each from psychology, the social sciences, and education, and 1 per- cent each from physics, engineering, and the professions. The biosciences as a doctorate field draw heavily from the undergraduate fields of chemis- try (15 percent) and agricultural sciences (8 per- cent) and less from others—3 percant each from medical sciences and education and 2 percent each from physics, engineering, and the humanities. Psychology draws a surprisingly high 10 percent from the humanities, 5 percent from the social sciences, 4 percent from education, 3 percent from the professions, and 2 percent each from the biosciences and mathematics. The social sci- ences draw heavily (13 percent) from the humani- ties, somewhat less so from the professions (7 percent), 4 percent from mathematics, 3 percent each from the agricultural sciences and educa- tion, and 2 percent each from engineering and psychology. The humanities draw 4 percent of their PhD's from the social sciences, an equal percentage from education, 3 percent from the professions, and not over 1 percent from any other field; 79 percent of the humanities doctorates had undergraduate training in the same field group. The professions, by contrast, are a very miscel- laneous set, and their undergraduate sources show it. The humanities contribute 22 percent; the social sciences, 13 percent; engineering, 8 percent; education, 4 percent; psychology, 3 percent; and mathematics, 2 percent. Education is also very broad in its undergraduate origins: humanities, 20 percent; social sciences, 8 per- cent; psychology, 6 percent; professions, 5 per- cent; mathematics, 4 percent; biosciences, 4 percent; chemistry, 2 percent; and agricultural and medical sciences, 1 percent each. THE GEOGRAPHY OF DOCTORATE ORIGINS The major change in the geography of doctorate production has been the rise of the South and Rocky Mountain States in the output of PhD's. In this section we look at these data from a different perspective—the regional interchanges between the baccalaureate and doctorate degrees and, going farther back, the regional inter- changes from the level of high school graduation to doctoral graduation. The map in Figure 47 shows the states in each region, and the accom- panying table (Table 25) shows the 1970 popula- tion in each region. One of the simpler ways of looking at the data of regional interchanges is to consider the ratio of the number of doctorate-bound baccalau- reates a region produces to the number of doc- toral degrees granted in that region. One may think of this ratio as a donor/receptor ratio, since all regions "give" students at one level to all other regions and "receive" students from all regions for graduate education. If this giving and receiving were equal, the ratio would be 1.00. If a region gives more than it receives

OCR for page 28
TABLE 25 DONOR-RECEPTOR RATIOS' AT TWO EDUCATIONAL LEVELS, BY SEX, FOR EACH U.S. REGION, 1960-1974 Region High School to PhD Men Women Total Baccalaureate to PhD Men Women Total New England 0.84 0.81 0.83 1.12 1.07 1.11 Middle Atlantic 1.40 1.24 1.37 1.16 1.07 1.15 East North Central 0.85 0.86 0.85 0.88 0.91 0.88 West North Central 1.22 1.31 1.23 1.21 1.27 1.22 South Atlantic 0.84 0.89 0.85 0.84 0.88 0.85 East South Central 1.26 1.40 1.28 1.28 1.32 1.29 West South Central 1.08 1.20 1.10 1.08 1.17 1.10 Mountain 0.83 0.72 0.82 0.91 0.75 0.89 Pacific 0.79 0.77 0.79 0.84 0.85 0.84 Donor regions are those with ratios over 1.00; receptor regions have ratios under 1.00. East South Central West South Central Mountain Pacific TOTAL US. Sates in Each Rej.on 1. New England: Maine, Vtrmont, New Hampshire, Massachusetts, Rhode Island, Connecticut 2. Middle Atlantic: New York, New Jersey. Pennsylvania 3. East North Central: Ohio, Indiana, Illinois, Michigan, Wisconsin 4. West North Central: Minnesota, Iowa, Missouri, North Dakota, South Dakota, Nebraska, Kansas 5. South Atlantic: Delaware, Maryland, D.C., Virginia, West Virginia, North Carolina, South Carolina, Georgia, Florida 6. East South Central: Kentucky, Tennessee, Alabame, Mississippi 7. West South Central: Arkansas, Louisiana, Oklahoma, Texas 8. Mountain: Montana, Wyoming, Colorado, New Mexico, Arizona, Utah 9. Pacific Washington, Oregon, California, Alaska, Hawaii (plus Puerto Rico and Panama Canal Zone) 1970 Population by Census Region (in thousands) New England Middle Atlantic East North Central West North Central South Atlantic 11,842 37.199 40.252 16.319 30.671 12.803 19,321 8.282 26^23 203,212 SOURCE: NRC, Commission on Human Resources FIGURE 47 The nine census regions of the United States.

OCR for page 28
69 TABLE 26 TIME CHANGES IN DONOR/RECEPTOR RATIOS AT TWO EDUCATIONAL LEVELS, 1960-1974 High School to PhD Baccalaureate to PhD 1960- 1965- 1970- 1960- 1965- 1970- Region 1964 1969 1974 1964 1969 1974 New England 0.77 0.82 0.88 1.01 1.13 1.15 Middle Atlantic 1.26 1.37 1.43 1.06 1.15 1.18 East North Central 0.81 0.87 0.86 0.84 0.89 0.89 West North Central 1.23 1.26 1.21 1.21 1.25 1.21 South Atlantic 0.93 0.84 0.84 0.93 0.84 0.84 East South Central 1.60 1.31 1.18 1.63 1.29 1.20 West South Central 1.21 1.12 1.05 1.22 1.11 1.05 Mountain 1.14 0.84 0.74 1.25 0.90 0.81 Pacific 0.76 0.74 0.83 0.81 0.79 0.89 SOURCE: NRC, Commission on Human Resources. from other regions, its ratio is higher than 1.00; if it grants more doctorates than it con- tributes to other regions at the undergraduate level, its ratio would be lower than 1.00. We can thus think of the regions with high ratios as donor regions and those with lower ratios as receptor regions. In these very simplified terms, the regions with older, well-established doctorate- granting institutions are the prime receptor regions. This group includes the East North Central States, the South Atlantic region, the Pacific Coast, and, for the most recent decade, the Mountain States. Prior to 1965, the Mountain States were in the donor category, but they have made a dramatic shift and are now in the receptor category. No other region has shifted across the balancing line of a 1.00 ratio, although the southern states have moved strongly in the same direction. Rather surprisingly, New England is in the donor category—apparently because its excellent undergraduate institutions attract many high school graduates from other regions, so that it donates more doctorate-bound baccalaureates than it graduates PhD's. At the high school to PhD interchange. New England exhibits a sharp contrast to its perfor- mance at the baccalaureate level. Because of its relatively small population, it produces fewer high school graduates that eventually attain the doctorate than it does either baccalaureates or doctorates. It is the only region that shifts from the receptor to the donor category between the high school and undergraduate levels of edu- cation. Tables 25 and 26 provide the information with respect to the relevant ratios. Table 25 shows the data for the entire 1960-1974 period, by sex, for both the high school/doctorate shifts and the baccalaureate/doctorate shifts. Table 26 shows the time changes, by 5-year cohorts, at both levels, for the combined total of both sexes. Tables 27 and 28 show all the regional interchanges for the entire 1960-1974 period. More detailed tables, by field, sex, and time period, are available from the Commission on Human Resources. Note that foreign areas are excluded in Tables 25 and 26 but given in Tables 27 and 28. Sex differences in the donor/receptor ratios are quite distinct although usually not as dra- matic as the changes over time. The patterns of sex differences are similar at the high school and baccalaureate levels, although the magnitude of the differences, and the range of the donor/ receptor ratios, is greater at the high school level than at the baccalaureate level. The regions in which the HS/PhD ratios and the BA/ PhD ratios are higher for men than for women are the Rocky Mountain States and the New England and the Middle Atlantic States. In the other five regions—the Pacific Coast, the South Atlantic States, and all the Central State re- gions, the ratios are higher for women than for men. That is, the tendency to "donate" rela- tively more men than women is stronger in the central regions and the Pacific Coast, while the East Coast and the Rocky Mountain States have a stronger tendency to "donate" women des- tined for the doctorate degree. This may be in part a result of field differences that have not been examined, since there are substantial sex and regional differences in the field mix at both the baccalaureate and doctorate levels, and they may be related in such a way as to produce the sex differences that have been noted in the donor/receptor ratios.

OCR for page 28

OCR for page 28
71 REGION OF BACCALAUREATE 100 50 OC o o o 0 w tr X o LU o z O HI rr i o ul O K LU O. 10 New England (NE) Middle Atlantic IMA) East North Central (ENC) West North Central (WNCI South Atlantic ISA) East South Central (ESC) West South Central (WSC) Mountain Statet Pacific & Insular (Pac) Foreign - MA ENC SA WSC (Mini Pac NE Geography has its "donors" too. each region is its own favorite source of baccalaureates for its eventual PhD's WNC ESC Mtn ENC MA ENC SA Pac MA Pac ENC ENC ENC ENC ENC P«r Nr- SA MA MA WSC ENC ENC NE SA WNC — SA Pac Pec Mtn WNC WNC MA SA Mtn - Pac SA NE MA WSC ESC WNC Pac MA NE NE MA WSC SA Pac ESC. Mtn WSC - Mtn WSC WNC MA SA WNC SA Mtn WNC WMC WSC NE WNC PK NE NE NE — Mtn Mtn Mtn Mtn WSC ESC WSC WSC ESC ESC ESC ESC ESC ESC SOURCE: NRC, Commission on Human Resources FIGURE 48 Graph of baccalaureate regional "donor" percentages. REGIONAL DONOR PERCENTAGE DIAGRAMS To provide a visual picture of the regional inter- changes. Figures 48, 49, 50, and 51 show the individual region-to-region percentage changes at both the baccalaureate-to-doctorate levels and the high-school-to-doctorate levels. At both levels, each region is considered from both the donor and receptor points of view; hence there are four figures in all. By examining these four figures (or the data of Tables 27 and 28) it is possible to develop a sense of the interregional interchanges that are occurring to move people from the high school and baccalaure- ate levels to the doctorate level. It should be noted, in examining Figures 48 through 51, that the vertical scale is logarithmic. This was done to bring into sharper focus the smaller

OCR for page 28
72 REGION OF HIGH SCHOOL GRADUATION 100 60 10 g 5 New Middle East North West North South Eest South West South Mountain Pacific US TOTAL England Atlantic IMAI Central (ENCI Central (WNCI Atlantic (SA) Central IESC) Central IWSCI State* & Insular IPac) Foreign INE) From high ichool to PhD the regional interchange* are heavier then from BA to PhD IMtn) — but have similar patterns — MA ENC SA WSC Pec — WNC ESC Mm NE ENC ENC MA ENC Pec MA SA Pec MA ENC ENC ENC ENC ENC ENC Pec SA NE SA PAC MA WSC ENC NE SA Pec SA WNC PK Mm WNC NE MA WNC Mtn MA Pec WNC SA WSC NE Pec - SA WSC MA ESC Pec MA WNC MA NE WSC Mtn NE 5A ESC WSC Mm WSC Pec WNC SA WNC SA WNC WNC NE WNC NE MA NE Mtn ESC Mtn Mm WSC Mtn Mtn NE WSC WSC WSC ESC ESC ESC ESC ESC ESC ESC SOURCE: NRC. Commiaion on Human FtetourcM FIGURE 49 Graph of high school regional "donor" percentages. percentages that characterize the interregional changes, in contrast to the "in-breeding" ratios (the diagonal data of Tables 27 and 28). Each region is, by a good margin, its own best source of doctorates—with the single exception of New England at the high school level. New England gets more doctorate-bound high school graduates from the nearby Middle Atlantic States than it does from its own high schools. We will begin an examination of Figure 48— the baccalaureate donor percentage diagram--with the column for New England. Here we see that New England contributes about 37 percent of its own doctorates, the smallest self-contribution figure for any of the regions. It contributes 21 percent of its BA's to the contiguous Middle Atlantic region, and 15 percent to the East North Central region. Next in order is the Pacific region, distant as it is geographically, closely followed by the nearby South Atlantic region. Far down—below 4 percent—are the other four regions, all more distant and with fewer vigorous doctoral institutions. In a similar manner the donor characteristics of the other regions may be examined. It is noteworthy that for each of the regions, its contribution to its own doctor-

OCR for page 28
73 REGION OF DOCTORATE 100 50 K o U at u, O O ui K o P 10 < z o CC o in b tv «/> O 5 o Ul cc z LU U cc 111 CL New Middle East North West North South East South West South Mountain Pacific England Atlantic Central Central Atlantic Central Central States & Insular (NE) (MA) (ENC) (WNC) (SA) (ESC) IWSC) (Mtn) (Pac) - Each region's PhD's earn their BA's — primarily in the same region - MA WSC — ENC WNC SA ESC Pac NE Mtn - - MA Foreign Foreign Foreign Foreign ENC Foreign MA SA WNC ENC ENC MA Foreign ENC WSC Foreign Pac ENC NE Foreign ENC WNC NE ENC Foreign WNC MA ~ Pac MA ESC ENC MA NE WSC Mtn NE WNC SA SA SA, Pac WSC MA r"^*p SA - Pac WSC WNC WNC MA WNC Pac WSC ESC Mtn Pac Pac SA - WNC NE SA Mtn NE SA - wsc Mtn ESC NE NE WSC Mm WSC. Mtn Mtn Pac ESC ESC Mtn ESC ESC SOURCE: NRC, Commission on Human Resources FIGURE SO PhD regional "receptor" percentages from each region of baccalaureate. ate production ranges somewhat above or below the 50 percent line but that no region contrib- utes more than 21 percent of its baccalaureates to any other single region. Typically, the interchanges that rank highest are between near- by regions but this is not always the case, par- ticularly with regard to the West Coast. Finally, to the right is a column for the total of all foreign regions of baccalaureate. The foreign regions, taken as a totality, contribute one--

OCR for page 28
74 REGION OF DOCTORATE 100 50 0 o o 05 I u— o o III K id z O o OJ ac u. O t- rr W a. 10 1 New Middle East North West North South East South West South Mountain Pacific U.S. TOTAL England (NE) Atlentic (MA) Central (ENC) Central (WNC) Atlantic (SA) Central (ESC) Central (WSC) States (Mtn) & Insular (Pac) - ENC WNC SA ESC i— WSC Mtn Pac MA ^ E xcept for New England (by a hair) each region's NE ^ F hD'j originate primarily in its own high schools MA - MA Foreign Foreign Foreign Foreign MA Foreign ENC Foreign SA Foreign WNC ENC ENC Foreign Pac Foreign MA ENC ENC WSC ENC - Foreign MA WNC — MA Pac — ENC WNC ENC MA WNC SA NE ENC WSC -SA Pac ESC MA ESC WSC WNC Mtn NE SA WSC Pac MF SA Pac NE WSC WNC - WNC WNC Mtn Pac Mtn SA NE ESC WSC SA WNC ESC SA Pac Pac Mtn NE WSC WSC Mtn NE - NE Mtn ESC Pac NE ESC WSC Mtn ESC Mtn Mtn ESC ESC SOURCE: NRC, Commission on Human Resources FIGURE 51 PhD regional "receptor" percentages from each region of high school. fourth of their number to the East North Central region, one-fifth to the Middle Atlantic, one- sixth to the Pacific region, and less than 10 percent to each of the other regions. Going back one educational level, we see, in Figure 49, the analogous contributions of each region of high school graduation to the several doctoral regions. New England contributes 35 percent of its doctorate-bound high school grad- uates to itself, 21 percent to the Middle Atlantic States, 16 percent to the East North Central States, and less than 10 percent to each other region. The pattern is very similar to the baccalaureate donor pattern but not exactly so. In comparing the regions at or near the bottom of the page, it may be noted that the

OCR for page 28
76 East South Central region, although it is typi- cally low except for the other southern regions, is never off the scale, as it is in the case of the baccalaureate origins. At the far right, beyond the foreign origin column, is a column for the total United States. What this column tells is the doctoral destinations for the entire U.S. doctorate-bound high school graduation popu- lation. The regions are, therefore, shown in terms of their relative outputs of doctorates of U.S. origin, which can be compared with their relative standing in output of doctorates from foreign secondary school sources, shown in the adjoining column. REGIONAL RECEPTOR PERCENTAGE DIAGRAMS The data of Table 25 can be examined in graphic form in Figures 49 and 50. Turning first to Figure 50, we see the pattern of baccalaureate receptor percentages—the percentage of each regions's PhD's that have been received from each of the regions of baccalaureate origin. New England receives 35 percent of its PhD's from New England undergraduate sources, 20 percent from the Middle Atlantic colleges and universi- ties, 15 percent from foreign sources, a bit over 10 percent from the East North Central region, and so on down the column. The Middle Atlantic States, shown in the second column, get half their doctorates from Middle Atlantic undergraduate schools, 15 percent from foreign sources, and less than 10 percent from any of the other regions. Each of the regions, as we scan across the diagram, is seen to be its own best undergraduate source, with the proportions ranging from about one-third to one-half of the region's doctorates. Foreign sources range downward from about 16 percent to about half of that for each of the regions of PhD. In no region except New England does another U.S. region contribute more than 15 percent to a region's PhD output. The pattern of secondary school sources for the various receptor regions, shown in Figure 51, is similar to that of Figure 50 but with some subtle yet pervasive differences. For example, the foreign area contributions, region by region, are higher at the high school than the baccalau- reate level, because some people with secondary education in foreign countries come to the United States for their undergraduate education. As mentioned earlier, the East South Central States never run off the bottom of the chart at the high school level, as they occasionally do at the baccalaureate level. The pattern of these differences suggests that a more intensive study than is possible in this book may well be re- warding. Such a more intensive examination, should scholars in this area be interested in pursuing it, could follow the movement, by sex and field, from high school to college to gradu- ate school and eventually on to employment. Many of the data necessary for such a study were pub- lished in the book Mobility of PhD's, published by the NAS in 1971; an update that takes into consideration the rather profound changes during the late 1960's and early 1970's—the period of "the new depression in higher education"—might be very revealing. It might be particularly revealing if it would take into account the edu- cational backgrounds from which the migrants and nonmigrants come, the nature of the jobs they eventually take, and some measures of career achievements. The necessary data for further studies of this nature, by university researchers or others, are available at cost from the DRF and Comprehensive Roster of the Commission on Human Resources.