Century of Doctorates: Data Analyses of Growth and Change : U.S. PhD'S--Their Numbers, Origins, Characteristics, and the Institutions From Which They Come : a Report to the National Science Foundation, to the National Endowment for the Humanities, and to (1978)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

1 Historical Trends HIGHLIGHTS • The number of PhD's awarded in the United States has approximately doubled in each decade over the past century. Quarter-century land- marks show that in 1900 the annual output was about 300; in 1925, about 1,200; in 1950, about 6,000; and in 1974, about 33,000. • World Wars I and II have produced the major fluctuations in the rate of growth of PhD production—first a dramatic drop, then an enor- nious rate of increase. The Great Depression of the 1930's had a less dramatic but nonetheless pervasive effect in lowering the rate of growth of PhD graduations. • The proportion of women among PhD's rose in this century from about 9 percent in 1900 to about 15 percent in the early 1920's, declined (except for World War II) to a low of 10 percent in the early 1950's, then rose sharply to over 20 percent in 1974. • The natural sciences claim about one-half of the PhD's among men; among women it is about one-fourth. Another one-fourth of the women are in education, which claims only about one-sixth of the men. • Proportions of PhD's in the various fields and field groups have varied over time; since 1970 the proportion in the natural sciences has diminished, and the proportion in education has increased markedly. • The number of living PhD's in the United States has increased since 1920 by a factor of 50, while the general population has approxi- mately doubled. • Among living PhD's, the fields of engineer- ing, education, and psychology—fields with a large "applied" component—have grown most rapidly; the other fields have grown at a more modest rate. GROWTH OF PhD AWARDS From the time the first earned PhD was granted in the United States—by Yale University in 1861—to the present day, the number of PhD's granted annually has increased at an average rate of about 7 percent per year, doubling every decade. The term PhD is used here to include equivalent third-1evel research degrees, such as ScD, EngD, and EdD, but excludes such profes- sional degrees as MD, DOS, DVM, or JD. The records of the U.S. Office of Education (USOE) for the years prior to 1920 are a bit uncertain and lacking in detail but are the best available. The data for the period since 1920 have been assembled from the Doctorate Records File (DRF) maintained by the Commission on Human Resources of the National Research Council (NRC). All data are in terms of calendar year unless other- wise noted. No attempt is made here to assess the quality of these degrees. We have simply counted the numbers as if each degree were equal to the others within the categories used here, such as field, sex, and cohort of graduation. The growth in PhD's can be envisioned in a number of ways—in terms of numbers of degrees granted, in terms of the fluctuations in the growth of numbers of degrees granted, and in terms of the resulting numbers of the PhD popu- lation. In this chapter, all of these approaches will be used, with a number of graphic techniques to aid in visualization of the data. A linear plot of the number of degrees granted annually over the past century averaged over 5-year intervals is given in Figure 1.

(A Ul K O t> o o . O c Ul m 35 30 25 20 15 10 The annual number of PhD's was about 300 at the turn of the century. 1.200 in 1925, 6.000 in 1950. 33,000 in the early 1970'! Annual Average by 5-Year Intervals 1890 1910 1930 YEAR 1950 1970 SOURCE: NRC, Commission on Human Resources FIGURE 1 Doctorates granted annually. While dramatic, this graph has a number of draw- backs from the standpoint of interpretation. The data cover a period in which the annual number of degrees increased a thousandfold. It is easier to visualize such an exponential growth process by plotting the data on a semi- logarithmic scale. This is done in Figure 2, which shows the average number of degrees granted per year for each 5-year period from 1875 through 1974. A straight line drawn through the "stair steps" of the graph depicts a steady 7 percent annual growth rate over this century. The deviations from this steady growth are in- formative, but one must allow for a greater degree of uncertainty of the data and the effects of small numbers in the years prior to the twen- tieth century. A slowing down is apparent for 15 years after 1895, and the year-by-year data of Table 1 show a particularly sharp decline during World War I. A growth spurt follows in the 1920's, then a slowing down during the years of the economic depression of the 1930's. Again, year-by-year data show a very sharp drop in PhD's granted during World War II and an upswing later that is even more dramatic than the huge step in Figure 2 at the beginning of the 1950's. Another slowing down appears after 1950; the growth of the "GI period" (about 1945-1950) was obviously not sustainable, and a secondary ef- fect of World War II appeared in the late 1950's. This was a lean period due to the inter- ruption and postponement of undergraduate edu- cation by the war; the gap moved on to the PhD level about 1957. Following this there is a steady increase through the 1960's, which ex- perienced the highest sustained growth in PhD output since the beginning of graduate education. The early 1970's show a sharp break in the growth curve. The output of PhD's, depicted graphically in Figures 1 and 2, is shown numerically in Table 1, which provides both annual data and 5-year sum- maries. As noted earlier, the data prior to 1920 are from the USOE, except for the years 1917 and 1919, which had to be filled in from NRC sources, since the USOE data became bien- nial after 1916. A third way of looking at PhD growth is shown in Figure 3, which depicts the 5-year summaries in PhD graduation numbers as succes- sive tree rings, each ring adding to the previ- ous number of doctorates granted. In Figure 3, the area of each new ring is proportional to the number of new degrees granted in the 5-year 50,000 10,000 5,000 § 1.000 8 O 500 m D 100 60 10 A logarithmic icalt of numbers of PhO's straightens out an exponential growth curve, revealing spurts and slow downs 1880 1900 1920 1940 1960 1960 YEAR SOURCE: NRC, Commission on Human Resources FIGURE 2 Doctorates granted annually (logarithmic icale).

TABLE 1 DOCTORATES GRANTED ANNUALLY BY U.S. UNIVERSITIES, 1875-1974," WITH 5-YEAR SUMMARIES (Calendar Year Data) Year PhD Total Year PhD Total Year PhD Total Year PhD Total 1875 23 1900 382 1925 1,206 1950 6,535 1876 31 1901 365 1926 1,441 1951 7,331 1877 39 1902 293 1927 1,540 1952 7,717 1878 32 1903 337 1928 1,632 1953 8,380 1879 36 1904 334 1929 1,917 1954 8,708 1875-1879 161 1900-1904 1,711 1925-1929 7,736 1950-1954 38,671 1880 54 1905 369 1930 2,075 1955 8,905 1881 37 1906 383 1931 2,344 1956 8,516 1882 46 1907 349 1932 2,400 1957 8,611 1883 50 1908 391 1933 2,462 1958 8,838 1884 66 1909 451 1934 2,696 1959 9,370 1880-1884 253 1905-1909 1,943 1930-1934 11,977 1955-1959 44,240 1885 77 1910 443 1935 2,529 1960 9,998 1886 84 1911 497 1936 2,713 1961 10,827 1887 77 1912 500 1937 2,752 1962 11,975 1888 140 1913 538 1938 2,754 1963 13,515 1889 124 1914 559 1939 2,950 1964 14,951 1885-1889 502 1910-1914 2,537 1935-1939 13,698 1960-1964 61,266 1890 149 1915 611 1940 3,277 1965 17,110 1891 187 1916 667 1941 3,484 1966 19,202 1892 190 1917 664 1942 3,404 1967 21,216 1893 212 1918 556 1943 2,592 1968 24,328 1894 279 1919 371 1944 1,967 1969 27,417 1890-1894 1,017 1915-1919 2,869 1940-1944 14,724 1965-1969 109,273 1895 272 1920 562 1945 1,634 1970 31,489 1896 271 1921 662 1946 1,990 1971 33,163 1897 319 1922 780 1947 2,951 1972 34,458 1898 324 1923 1,062 1948 3,940 1973 33,472 1899 345 1924 1,133 1949 5,389 1974 33,165 1895-1899 1,531 1920-1924 4,199 1945-1949 15,904 1970-1974 165,747 Preliminary data received too late for further analysis indicate that in 1975 there were 33,146 PhD's granted; 33,200 were estimated for 1976; 32,000 for 1977. SOURCE: NRC, Commission on Human Resources. period so that the total area shows cumulative numbers of degrees. This provides a beginning for consideration of the PhD population, as distinct from graduation numbers, a topic that is taken up in more detail later in this chapter. Most of the data available with respect to doctorate output and the characteristics of PhD's comes from the period since 1920, which marks the beginning of the DRF of the Commission on Human Resources of the NRC. Although data collection for the DRF began only in 1946, it was possible to go back to the universities and obtain graduation records, permitting the beginning of a name file, with individual data on each graduate. A decade later a further step was taken, with the initiation of a ques- tionnaire, the Survey of Earned Doctorates, which was filled out by each graduate and for- warded to the NRC. This permitted more infor- mation and more accurate information with respect to the graduate, including his or her own statement as to the fields of specializa- tion at the time of all degrees earned and where and when the degrees were earned. This in turn permitted study of baccalaureate-to- doctorate time lapse, the switching of fields between baccalaureate and doctorate, geographic migration, and a number of other topics de- scribed in later chapters. Growth of PhD output during the 1920-1974 period is depicted graphically on a linear plot in Figure 4. The data here are 3-year moving

Each 5-vear per iod adds a new ring of growth to the previous number., the edded eree ii proportional to the number ol new PhD's SOURCE: NRC. Commission on Human Resources FIGURE 3 Growth in doctorates depicted as tree rings. 30 g f U. O I < S3 o 20 Smoothing the growth curve accentuates the longer-term changes in growth; the 1970's saw a sharp break in the curve 1920 1930 1940 1950 YEAR 1960 1970 SOURCE: NRC, Commission on Human Resources FIGURE 4 Growth in doctorates since 1920.

+40 +30 +20 | +10 S o g -10 u rr uj -20 The depression of the 1930's and World War \ \ slowed growth rates; the postwar spurt could not be sustained 3-Year Moving Average of Increments Long-Term Average Growth Rate 1+7%) 1920 1930 1940 1950 YEAR 1960 1970 SOURCE: NRC, Commission on Human Resources FIGURE 5 Growth increments in doctorates granted. +50 +40 S +30 K U +20 +10 UJ o oc UJ -20 -30 The science fields all respond to major in- fluences but with slightly different patterns of growth rates Life Sciences Behavioral Sciences Engineering. Meth, and Physical Sciences 1930 1940 1950 1960 1970 YEAR SOURCE: NRC, Commission on Human Resources FIGURE 6 Growth increments in doctorates granted in three science fields. averages, which show chronological changes more faithfully than the 5-year summary data. Such averages iron out the year-to-year changes that are to a certain extent random, depending on minor factors such as universities' policies with respect to when graduations occur or the month in which all requirements are finally met. In Figure 4, the flattening of the growth curve during the depression of the 1930's is shown, as is the deep decline in output during World War II. The long steep rise of the 1960's is followed by a sharp change in the 1970's, in- cluding an acutal drop in output for the first time since 1957. GROWTH INCREMENTS Changes in rate of output of PhD's are more readily visualized in a graph of percentage increments or decrements. These data, calcu- lated on an annual basis, are somewhat unstable and are best viewed after smoothing by means of a moving average. Figure 5 shows such a graph for the period from 1920 through 1974. Here the changes due to wars become dramatically apparent and the depression of the 1930's shows a gradual decline. The drop during the 1970's, following the prosperous 1960's, is even more evident than in the linear output graph of Figure 4. Figure 5 shows total output figures; some breakdown by fields may be useful in consider- ing the possible causes and consequences of the changes that have occurred. Figure 6 shows the analogous curves for three field groups:

10 + 40 -30 ui o +20 3 +10 1 ° 5.10 re ui C -20 Before World War II, education giuwlh rates were more volatile, in part bacauM of tha innabilitiat of imall number s 1920 1930 1940 1950 1960 1970 YEAR SOURCE: NRC. Commission on Human Resource FIGURE 7 Growth increments in doctorates granted in nonscience fields. +15 +10 UJ 3 UJ K o z u c UJ o, -6 Are th« EMP fields a "leading indicator"? The| trends are suggestive. . . . I I I I I I I I I 1952 1955 1960 I I I I I I I I I I I \l 1965 1970 1973 YEAR SOURCE: NRC, Commission on Human Resource) FIGURE 8 Increments in doctorates granted in three science fields 1952-1974 (moving averages). (1) EMP fields, (2) life sciences, and (3) be- havioral sciences. Figure 7 shows the same kind of data for the remaining major field groups: humanities, professions, and education. Data on growth by field by year, with 5-year summaries, are given in Table 2. The numerical data for the series of increment graphs are given in Table 3, for those who wish to examine the data in more detail. The most intriguing data, however, relate to the performance of the science fields for the most recent period, as shown in Figure 8.

11 z Ul LU c u o 2 Ul u K +40 +30 +20 + 10 -f? -10 -20 - PhD growth rates for women com- pared to those for men were affected earlier and more strongly by the depression and less by World War 11 and have been recently stimulated by the women's movement I I 1920 1930 1960 1940 1950 YEAR SOURCE: NRC, Commission on Human Resources FIGURE 9 Growth increments in doctorates by MX. 1970 the three science field groups shown in Figure 8 is striking. The EMP fields behave like a "leading indicator"—to borrow a term from the jargon of economics. The fluctuations in the life sciences output are closer to the general average of all PhD fields, while the behavioral sciences show a lag, moving downward, upward, and downward again later than the other fields. These variations cannot be accounted for di- rectly from the data at hand; a number of studies have been made and are being made of the determinants of doctorate output. As the results are as yet inconclusive, no attempt will be made here to account for the rather striking curves of Figure 8. One factor affecting the time trends in out- put of PhD's that is evident in the preceding graphs is the economic climate. Another is the effect of wars. These two influences affect the two sexes differently, and the result of these, as well as other influences, is shown in Figure 9, which shows the incremental changes since 1920 for men and for women separately. (The graph here is not a 3-year moving average, but a 4-year center-weighted moving average, which is somewhat more stable, though slightly less sensitive. This center-weighted average doubles the data for the 2 middle years and divides the sum by 6. It was chosen to iron out the random fluctuations that occur with small numbers, as, for example, with women in the earlier years of this period.) It is clear from Figure 9 that the effect of World War II and its aftermath was greater for men than for women, as expected. The figure also suggests that the earning of doctorates by women is highly sensitive to the economic climate, as shown in the 1930's; during the depression the curve for women dropped earlier and more steeply than did that for men; in the most recent period, the drop in increment started earlier for women in the "academic depression," which began in 1968. It was not so severe as the drop in the curve for men for a number of reasons, probably the principal one being the different "field

12 TABLE 2A DOCTORATES AWARDED ANNUALLY IN ENGINEERING, MATHEMATICS, AND NATURAL SCIENCES, 1920-1974, WITH 5-YEAR SUMMARIES 1920 1921 1922 1923 1924 562 662 ill 21 129 19 7 10 15 199 38 34 42 H 109 91 12 \l 27 45 32 IB II! ill 12 22 174 217 15 17 34 29 199 111 27 19 28 34 780 1062 1133 55 140 185 224 249 333 373 69 102 100 157 242 216 11 40 285 330 14 14 ti 169 150 44 589 TOTAL 1920-24 4199 245 751 139 1135 114 60 1309 231 395 626 120 136 882 2191 111! 1206 1441 1540 1632 1917 51 211 27 269 28 16 333 456 427 69 110 179 30 33 24 20 39 36 245 1 87 81 252 217 255 251 Jf 31 48 381 343 48 27 77 120 121 155 164 197 224 29 259 290 33 51 41 103 II 68 5*05- .si in 60 TOTAL 1925-29 7736 411 1186 193 1790 237 161 2195 459 670 1125 146 223 1494 3689 1930 1931 1932 2075 ISO'S itti m 302 334 328 382 419 66 474 76 82 74 7} 91 64 614 103 130 132 153 175 169 225 201 205 246 III 46 52 40 44 64 61 62 379 993 1107 1096 1233 1393 115 42 55 74 68 489 498 589 607 67 68 92 119 638 640 756 817 333 83 75 91 469 456 477 Hi2 III 358 421 976 TOTAL 1930-34 11977 991 1761 309 2657 398 410 3465 693 1046 1739 246 372 2357 5822 ill! 2529 133 J''S 11 564 R 111 750 126 150 163 218 242 233 275 254 359 425 417 476 509 47 80 486 614 1236 III 7'H 70 798 885 772 852 45 26 51 36 1? 68 69 sill 1938 1939 2754 2950 156 160 409 461 70 635 690 62 93 98 75 69 258 267 1367 1466 62 TOTAL 1935-39 13698 742 2189 323 3254 3M 423 4057 •99 1287 2186 209 336 2727 6784 1940 1941 1942 1943 1944 3277 934 59 64 66 43 19 737 102 96 76 44 44 m W 271 228 175 l?l 296 218 130 47 53 ii 705 Wl 3484 3404 2592 1967 "-* 158 647 590 511 475 890 814 686 559 663 729 564 391 132 69 98 53 69 988 783 567 446 305 60 43 39 ,02 47 Hi? 668 1059 TOTAL 1940-44 14724 671 2757 251 3686 362 445 4493 1178 1221 2399 242 411 3052 7545 1945 1634 4} 290 326 23 356 36 if 460 120 91 97 147 31 8 |?J 762 909 1386 1986 2907 1946 1947 1948 1949 73 38 59 70 111 437 625 18 104 120 258 451 595 860 1292 1964 146 238 \\ ,8! 526 3940 5389 230 314 421 615 935 HZ 218 288 258 314 386 913 674 86 183 694 943 TOTAL 1945-49 15904 •09 2587 30« 3700 470 1001 5171 •63 1202 2065 250 464 2779 7950 m 422 1052 130 1604 1683 1732 1697 1702 1 469 24741 324 lit 1148 1135 iiii 33H? 111! 1034 1063 1008 1018 148 149 586 UK 2512 404 496 599 596 ill 150 JS.*' Hog 1*1 160 563 539 332 370 1655 4167 TOTAL 1950-54 38671 2489 5175 754 8418 1056 2756 12230 2255 2570 4825 570 1536 6931 19161 1955 1956 1957 1958 1959 111 9370 8905 m 1013 119 157 f|7 199 239 1704 649 2596 575 540 487 513 511 534 1115 164 368 356 345 325 362 1647 1524 1632 1584 1627 4243 ,o\8i 1623 1692 1661 1835 579 589 658 712 2430 2537 2556 2848 488 619 609 975 1132 193 155 139 199 5954 4169 4140 4475 464 504 523 958 1077 237 301 976 1120 1110 TOTAL 1955-59 44240 2487 5070 991 8515 1265 3187 12967 2867 2585 5452 •06 1756 8014 20981 1960 1961 1962 1963 1964 9998 10827 11975 13515 14951 974 Up 7 38 1192 1356 1370 251 1932 289 825 1006 1288 1453 1804 3046 636 559 1195 146 172 429 isll 4816 601 767 829 936 261 257 326 325 2000 2216 2511 2631 362 411 537 620 3368 3915 4501 5055 697 723 509 621 1206 1344 1480 1672 208 209 278 433 510 472 537 5179 5977 6662 7542 ,57 ffl TOTAL 1960-64 61266 3707 6163 1420 11290 2219 6376 19885 3826 3071 6897 1013 2381 10291 30176 1965 17110 19202 21216 24328 27417 1065 1187 1361 1447 1575 1480 1712 1793 1824 2129 395 734 1074 • 19 J19 567 2775 8635 il 434 11 803 849 1046 1131 5860 6573 7042 7783 8694 1230 1367 1628 1740 881 1889 2111 2358 2810 3052 314 355 409 452 604 647 689 892 3029 3360 3908 4396 9602 10402 11691 13090 Jit 499 4203 3360 991 H82 1312 TOTAL 1965-69 109273 6635 8938 2227 1 7800 4563 13589 35952 7039 5181 12220 1849 3399 17468 53420 1970 1971 1972 1973 1974 31489 33163 34458 33472 33165 1715 2284 2248 2007 1631 1800 534 4533 if!! 3603 9418 1823 1504 1569 1556 1427 1396 3327 3479 3527 3247 3203 544 1012 4883 29 5224 69 5264 105 4937 116 5013 14301 1743 1697 1412 1360 564 636 575 574 4555 4340 3818 3734 1215 1155 3654 3493 3259 3039 9483 9174 8292 7928 1890 1971 1820 1807 607 604 583 611 1109 1064 1002 1083 14707 14438 13229 12941 TOTAL 1970-74 165747 7927 10170 2883 20980 6267 17048 44295 9311 7472 16783 2949 5270 319 25321 69616 GRAND TOTAL 487435 26717 46747 9761 83225 17331 45463146019 29617 26700 56317 8396 16284 319 81316 227335 SOURCE : NRC, Coi mil ssion on Human Re: sources

13 TABLE 2B DOCTORATES AWARDED ANNUALLY IN THE BEHAVIORAL SCIENCES, TOTAL OF ALL SCIENCES, AND NONSCIENCE FIELDS, WITH 5-YEAR SUMMARIES £ /.? z /J1 // / // /! */ /y /s // / // /? // // /•? * /* // Hi1? it M 11 if | 87 376 8 ii 119 t id 14 111 **3 507 150 179 1922 33 101 56 34 45 44 32 59 1,11 65 55 40 52 js ii 8 5 150 161 725 750 61 ** 60 57 48 65 69 46 222 228 *5 68 52 102 2 337 1 383 TOTAL 1920-24 21T 185 78 104 26 610 2801 238 188 242 230 898 181 310 9 1398 1925 7^ 64 29 28 13 205 783 63 55 57 60 235 56 128 * *23 1926 81 26 33 13 227 942 71 71 55 76 27 3 64 161 1 499 1927 76 91 29 Ii If 258 975 88 63 64 88 303 88 170 * 565 1928 1929 84 122 85 103 25 37 ii III IB? 94 70 107 69 68 94 382 77 173 7 572 18 696 TOTAL 1925-29 427 424 168 195 78 1292 4981 *23 328 338 419 1508 370 8*3 34 2755 1930 1931 10,1 107 5*0 33 57 ii in 1306 128 96 118 108 95 96 124 415 75 268 11 769 11 874 119 1470 115 453 107 303 1932 105 122 57 58 20 362 1458 123 129 137 504 124 309 5 942 1933 1934 92 128 108 113 52 52 68 65 27 24 347 382 1580 1775 1*8 11* 1*8 137 140 109 \\\ 103 261 103 280 7 882 13 921 166 TOTAL 1930-34 545 569 256 281 116 1767 7589 665 584 641 518 2408 512 1421 47 *388 90 4 341 iS 156 136 135 1** 1** 161 159 159 174 547 134 254 1935 112 52 1577 81 98 80 84 17 952 1936 1937 1938 11* 112 103 108 125 f! i| 10 338 369 368 185 169 172 562 554 574 103 355 80 358 83 362 27 1047 * 996 1019 116 61 10 1939 117 112 75 60 14 378 184* 177 173 164 106 620 109 377 1106 TOTAL 1935-39 571 538 318 295 72 1794 8578 771 773 864 449 2857 509 1706 *8 5120 1940 129 125 73 82 25 434 2085 167 174 180 107 628 94 470 1192 1941 (13 158 91 72 13 447 2218 182 189 178 127 676 111 478 1 1266 1942 126 138 77 70 13 424 2141 168 177 150 126 621 1*8 493 1 1263 1943 194* 92 68 82 61 58 41 50 36 13 14 295 220 M 122 124 115 70 81 66 442 270 105 *02 103 31* 1 950 60 74 1 688 TOTAL 1940-** 528 564 340 310 78 1820 9365 699 738 693 50T 2637 561 2157 * 5359 1945 1946 65 82 59 B 28 39 5 9 m 951 71 74 119 112 K 62 79 279 380 108 294 I iff! 84 1183 79 347 1947 1948 1949 122 190 266 135 1*3 183 92 82 118 tsf tj 427 Bl 1813 2529 3662 169 165 146 167 224 179 ift lii 565 116 455 142 666 173 84* 41 Hi TOTAL 1945-49 725 604 384 384 91 2188 10138 729 697 551 558 2535 618 2606 7 5766 1950 1951 360 243 168 166 41 978 1195 4344 274 235 211 is? 933 219 1038 250 1110 1 21|1 488 299 189 165 54 4876 339 298 201 1095 1952 581 313 178 157 37 1266 5129 298 262 180 286 1026 247 1315 2588 656 IIS 164 186 58 50 1403 1503 M 350 333 364 34* 202 216 1223 1271 242 1425 260 1507 3 2893 3038 TOTAL 1950-54 2752 1516 999 838 240 6345 25506 1625 1472 1010 1*41 5548 1218 6395 * 13165 1955 I 800 734 379 229 213 53 88 55 76 62 1604 333 327 267 347 lit 340 310 361 391 395 1216 1145 1270 270 1572 275 1638 334 1384 343 1503 356 154* 3058 3 3061 3 2991 If? lo? 220 175 235 242 2*8 15?l 1958 1959 320 338 IK i^? 340 369 31* 310 351 346 IS! till i IB? 230 6147 230 TOTAL 1955-59 3650 1654 1101 1118 334 7857 28838 1605 1699 1074 1797 6175 1578 7641 8 15402 US? III 376 P* 80 64 81 till 6500 7036 7905 8869 9787 364 368 206 531 503 504 569 602 1*69 390 1632 440 1801 43* 1951 547 2296 586 2494 7 3498 7 3791 8 *070 12 *646 17 5164 nn If! 434 4M 497 534 280 m 1928 2207 2245 in ii! SB a? in If? 1964 968 309 296 318 346 300 337 1791 2067 TOTAL 1960-64 4443 2249 1345 1*49 435 9921 40097 2101 2338 1349 2709 8497 24*7 10174 51 21169 !| 18( 572 111 322 413 *92 569 410 97 2473 11108 12395 13700 15426 72256 7?S 434 697 I 654 2892 804 3229 853 3659 67 *S?! 420 559 578 107 156 260 2793 702 847 820 995 *89 556 672 78* 832 1606 617 591 I??? 1031 *333 1969 1843 779 401 4231 17321 934 1035 682 1073 372* 1063 5086 223 10096 TOTAL 1965-69 7058 3480 2413 2558 1021 16530 69950 3807 *320 2833 *352 15312 **05 19199 *07 39323 1970 1 2119 842 864 991 901 785 883 904 1017 1031 HI l\l 38? iSl'i. 092 1227 796 m 111, *1*i 1*O* 6305 1415 6898 1627 7318 1425 7331 1*86 7219 ill (jit} 108 13891 2181 2386 2512 2741 925 673 741 5879 6087 6333 fl 1*3 1285 236 1*13 886 5049 5365 5078 90* 800 19274 969 TOTAL 1970-74 11939 4502 *620 *310 3389 28760 98376 5893 6668 *61* 7182 24357 7357 35071 586 67371 GRAND TOTAL 32855 16285 12022 11842 5880 78884 306219 18556 19805 14209 20162 72732 19756 87523 1205181216 SOURCE: NRC, Commission on Human Resources.

14 TABLE 3 THREE-YEAR MOVING AVERAGES OF ANNUAL PhD GROWTH INCREMENTS, 1920-1974, BY FIELD AND TOTAL Mathe- Phys- Che«- Earth Sci- Engl- TOTAL, EMP TOTAL, Life Sci- Psy- Social Sci- TOTAL, Behav- ioral Sci- TOTAL Sci- Human- Pro- fes- Educa- GRAND c ogy tion TOTAL 1921* -3.9 34.0 37.2 20.2 46.5 26.8 8.7 0.7 20.2 9.3 16.1 23.2 41.5 23.8 17.8 1922 30.7 25.7 35.5 40.7 28.7 29.1 23.8 30.9 22.4 22.4 25.1 23.2 41.2 21.0 23.9 1923 32.9 20.3 21.7 58.4 14.4 23.6 20.5 32.4 10.8 15.6 20.3 15.0 16.8 48.0 20.2 1924 27.3 -i.a 15.8 17.7 2.5 11.7 18.9 10.9 26.0 27.7 16.9 9.9 21.3 30.3 16.4 1925 17.8 18.7 11.6 9.0 27.7 12.7 2.9 6.0 21.8 15.1 9.4 7.1 12.5 33.8 10.9 1926 24.8 15.3 -0.1 8.0 35.1 6.6 10.3 12.0 19.9 17.2 9.4 10.3 19.8 19.0 10.9 1927 20.0 27.0 7.7 10.5 48.5 13.8 10.3 5.8 9.4 8.0 10.8 10.4 13.1 21.8 10.8 1928 16.9 4.2 0.7 10.3 19.0 3.7 12.6 19.5 14.3 15.8 9.1 13.4 11.8 9.8 10.1 1929 18.7 9.6 12.1 20.4 30.3 13.0 9.5 12.8 6.4 8.2 10.3 11.7 9.5 16.9 10.6 1930 27.2 6.O 9.8 18.6 13.7 10.7 12.9 15.3 13.0 13.6 15.9 13.2 14.0 20.7 12.9 1931 3.3 5.9 9.7 10.7 20.8 1.6 7.7 -3.7 5.2 2.1 10.6 8.6 15.9 14.0 7.9 1932 -0.2 a.o 8.4 9.7 13.8 7.4 8.5 -2.1 6.5 3.9 11.1 7.0 14.5 -0.1 6.0 1933 4.3 3.6 7.8 19.1 22.0 8.8 7.5 5.0 1.4 1.9 6.6 5.2 -0.3 -2.1 4.8 19 34 1935 !• 7 5.4 4*4 6.0 7 . 8 19 . 3 -4.8 6 . 0 } . 2 3.5 4.4 2.3 -6. 3 12.7 1.9 3.5 1.7 1.4 -1.1 2.1 4.8 9.5 -4.1 -0.5 2.2 4.4 1936 -6.3 7.8 7.7 -6.4 -1.2 3.0 -3.9 -4.2 0.9 -0.8 -0.1 1.1 -5.1 10.6 0.8 1937 -5.8 5.6 S.S 4.4 -6.8 l.S 7.4 1.2 1.5 2.7 3.3 0.7 -13.9 14.7 3.O 1938 10.4 S.2 3.1 -1.9 2.8 2.8 5.5 0.9 5.5 3.9 3.5 1.9 4.3 2.1 2.9 1939 14.9 -2.3 3.2 4.5 7.9 2.9 12.2 4.9 6.2 5.7 6.1 4.1 7.1 10.1 6.1 1940 18.0 S.6 16.6 -2.6 20.4 12.9 4.0 -0.4 10.0 6.8 8.6 5.6 11.9 10.1 8.2 1941 -5.6 0.7 8.8 2.3 16.5 5.6 6.3 3.1 5.2 4.2 5.1 0.1 12.5 9.8 5.0 1942 -23.2 -1.3 -0.3 -7.7 -17.2 -4.9 -6.2 -9.3 -11.1 -10.8 -6.8 -9.7 7.4 -4.8 -6.6 1943 -21.7 -26.3 -9.1 -29.9 -14.3 -15.6 -14.7 -13.9 -22.8 -20.4 -16.3 -25.4 0.8 -12.3 -16.9 1944 -19.5 -33.5 -20.0 -21.7 -6.2 -22.3 -25.3 -19.2 -25.1 -23.3 -23.7 -21.8 -9.0 -15.9 -21.8 1945 9.9 -4.9 -11.3 10.2 26.7 -5.6 -16.1 -1.4 4.0 2.0 -7.8 2.1 -7.7 -2.9 -6.2 1946 48.3 44.9 1.9 51.2 24.0 15.0 16.6 23.5 32.1 29.1 17.4 29.8 7.9 13.6 18.0 1947 53.5 74.3 28.9 45.2 61.1 40.8 34.0 41.1 42.9 41.8 39.0 30.8 13.8 32.3 34.3 1948 47.1 65.3 43.1 51.8 68.8 49.7 44.8 49.7 37.6 41.2 45.8 22.6 30.0 34.7 39.8 1949 15.7 43.1 36.2 32.6 65.2 38.6 28.6 43.7 26.7 31.9 34.3 18.5 23.6 32.3 30.4 1950 20.6 31.2 21.8 33.9 34.9 26.5 21.1 39.5 26.7 31.3 25.2 22.4 21.2 18.9 23.9 1951 11.6 18.1 4.2 7.3 8.7 8.2 12.8 28.5 12.5 18.7 12.0 14.2 13.0 15.9 12.6 1952 8.6 7.6 -1.3 27.2 7.1 3.6 12.7 22.5 6.9 13.0 8.1 10.1 3.7 11.4 8.7 1953 6.5 1.5 -0.5 21.2 -1.3 0.5 11.3 11.1 6.0 7.9 5.2 5.7 1.4 10.8 5.9 1954 6.2 -0.6 -1.6 25.1 4.7 1.2 6.9 8.3 8.4 8.3 4.5 6.2 3.1 6.2 4.9 1955 0.8 -2.4 -0.9 -1.5 1.2 -0.7 -1.4 -0.8 5.4 2.5 -0.4 -2.1 4.4 4.7 0.6 1956 1.5 -4.0 0.8 6.1 2.1 0.5 -0.3 3.2 -3.7 -1.0 -0.5 0.1 9.1 -2.3 -0.3 1957 -0.4 -0.3 -1.7 4.2 0.8 -0.4 -1.1 2.5 0.3 0.8 -0.7 0.5 8.7 -0.9 -0.2 1958 10.6 2.7 3.5 14.5 7.2 5.5 2.3 8.4 0.8 3.9 4.1 S.I 9.3 -1.4 3.2 1959 5.2 7.4 2.4 10.4 11.9 6.4 2.9 1.9 8.3 5.2 5.0 S.3 5.4 5.7 5.1 1960 16.1 6.1 6.0 9.6 15.3 9.7 4.6 4.0 5.2 4.5 6.8 8.1 8.8 6.4 7.0 1961 11.6 14.0 3.4 3.1 21.9 11.3 8.3 3.9 6.1 4.9 8.8 7.1 10.8 8.0 8.5 1962 23.2 13.5 7.1 9.8 20.9 13.9 7.0 9.1 10.0 9.5 10.9 6.9 12.0 12.1 10.6 1963 19.9 16.2 6.S 8.3 21.7 14.5 11.3 4.1 8.9 6.7 11.7 10.4 10.0 11.4 11.3 1964 21.5 11.6 7.6 16.0 19.4 14.4 10.5 6.6 10.6 8.8 12.0 14.6 10.5 14.2 12.7 1965 14.4 12.7 8.2 10.4 19.0 13.5 12.0 6.2 9.9 8.3 13.2 15.5 13.9 12.1 12.4 1966 11.2 13.3 9.5 9.2 13.4 11.7 10.6 12.4 14.7 13.7 13.2 12.6 13.5 13.6 12.4 1967 12.8 10.8 7.4 7.1 11.0 9.9 12.1 14.5 15.0 14.8 12.9 12.7 16.7 14.4 12.5 1968 12.3 9.9 7.7 5.0 11.4 9.8 13.2 16.6 13.7 14.9 11.8 11.5 10.0 16.4 12.6 1969 14.9 8.0 8.6 8.6 11.2 10.2 13.3 15.6 14.8 15.1 12.1 13.9 18.7 19.9 14.1 1970 7.0 6.4 7.5 S.4 7.1 6.9 10.2 10.8 15.4 13.4 9.4 9.3 12.0 16.9 10.9 1971 6.0 2.6 -1.7 8.1 1.4 1.9 6.2 8.9 13.7 11.6 S.S 10.0 16.0 13.1 8.0 1972 -1.8 -5.9 -7.0 2.9 -3.2 -4.1 0.4 5.7 7.3 6.6 0.1 7.2 1.2 5.2 2.1 1973* -2.4 -9.7 -9.7 1.6 -5.5 -6.5 -2.7 7.3 4.7 5.8 -1.4 8.6 l.S 1.2 o.s *Data for 1921 and 1973 are 2-year averages. SOURCE: NRC, Commission on Human Resources. mix" of men and women PhD's. Men are concen- trated more heavily in the sciences, particularly the physical sciences and engineering; women are concentrated more heavily in education, which has shown a remarkable increase in recent years. In addition, of course, the women's movement has in recent years been an important factor in higher education and advanced training. All these factors, as well as others, have kept the output of women PhD's at a high level. BACCALAUREATE DEGREES One of the basic factors involved in numbers of PhD's, quite obviously, is number of baccalaureate- level graduates. The trend in these degrees is shown in Figure 10, the figures for which come from the USOE. (For the period prior to 1961, the USOE data are for "baccalaureate and first professional" degrees; after 1961, the two degree types are separated. In Figure 10, a correction

1,000 500 100 50 10 Al the baccalaureate level, men art no longer the overwhelming proportion they once were (or were during the "Gl period") _ — — Women 7 •""••• Both Sexes Combii i i 1 I 1 1 1 1 1 1 1880 1900 1920 1940 1960 19 YEAR SOURCE NRC, Commission on Human Resources FIGURE 10 U.S. baccalaureates conferred annually. was introduced for the period of the 1950's; before that time the number of first professional degrees is too small to warrant a correction in the graphic display; the shape of the curve is not changed in any case.) In Figure 10, it is apparent that the curve for baccalaureate degrees granted to women is converging with that for men; this is an obvious source of influence for the corresponding but weaker tendency, somewhat later, at the doctorate level. Because BA-PhD time lapse varies by field and by time period, and because people switch fields between the baccalaureate and doctorate, it is not possible to demonstrate a close linkage between baccalau- reate output in a given period and PhD output at some later time. General trends only are shown in Figure 10; their significance may well be very important a generation later, as indicated in Chapter 2; no more definitive interpretation will be attempted here. Another factor frequently invoked to help to account for the changes in numbers of doctorates granted is that of financial support to research and development. There have been a number of attempts to relate such support to output in particular fields, as, for example, the biomed- ical sciences, but there is no real consensus on the importance and timing of the effects in variations in federal support for research. There are a number of reasons why the impact is neither immediate, direct, nor unambiguous. One is the differing impact of expenditures for basic research as distinct from development. A much higher proportion of basic research funds go to universities, as compared to development funds, in which the business and industry sector participates more heavily. Another reason why funds for research do not have an unambiguous effect is that they go, in an undetermined pro- portion, for salary of the principal investigator, equipment expenses, overhead, etc., and in some other proportion for the support of training of research personnel who also participate in the research. Figure 11, here reproduced from a National Science Foundation (NSF) report (NSF 77-311), depicts graphically the changes in federal obligations to universities and colleges over the period FY 1963-1975. The top graph shows total dollars, interpreted also 'in terms of constant 1972 dollars, using the GNP deflator. The bottom graph shows a breakout of the current dollar amounts into several categories. Figure 12, also from the NSF (NSF 76-310), shows the trends in funding, both federal and nonfederal, from 1953 through 1976 (the last 2 years esti- mated) . In both Figure 11 and Figure 12, whether current dollars or constant dollars are concerned, the long upward trend in federal sup- port ceased in 1967, and a decline, in constant dollar terms, set in. During the 1970's, the trends have been mixed, in constant dollar terms, with little net change in federal obligations to universities and colleges but a net drop in total federal funds for research and development (RSD), taken up in part by increases in nonfederal sources, as shown in Figure 12. A factor that cannot be shown by either of these charts is the fact that universities have their homeostatic mechanisms for adjusting to varying kinds and amounts of financial support. Historians have discovered evidence for such ad- justments as far back as the early 1800's, in the correspondence of Thomas Jefferson, concerned with support for the University of Virginia in its early days. Federal support for science, for example, may result in shifts of support from other sources toward the nonscience fields; each university finds its own means for maintain- ing balance despite fluctuations in "soft money" from federal sources. The effect of federal funds, therefore, while important, is diffuse. No doubt many students felt that, even though they had scant prospects of a typical academic job, nevertheless their prospects were better after attaining the doctorate than before, and they therefore persisted despite diminishing prospects in the faculty job market. Examination of these factors in student decision making and institutional adjustments, interesting as they are, cannot be further pursued in this report.

16 £ 5 Conslant 1972 Dollars* O 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 H 2 BY TYPE OF PROGRAM Academic Science Obligations, All Categories 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 FISCAL YEAR 'Based on GNP implicit price deflator SOURCE: National Science Foundation FIGURE 11 Federal obligations to universities and colleges, FY 1963-1974; growth by type of program. Average Annual Rale of Change Current Constant* Year Total Non- Federal federal Total Non- Federal federal 19531961 13.7% 163% 10.0% 11.4% 14.0% 7.8% 1961-1967 8.3 7.7 9.6 6.0 5.3 7.2 1967-1976 5.7 3.8 8.3 -0.3 -2.2 2.1 1953 "Bated on the GNP implicit price deflator SOURCE: National Science Foundation 1973 75 76 (est) FIGURE 12 R&D funding trends: 1953-1976.

17 THE ROLE OF WOMEN The fluctuations in the growth of PhD output and the differing trends of the growth increments for the male and female segments of the PhD grad- uating classes have been mentioned. This aspect of doctorate production needs more attention (since the changes shown are in part a function of the changing "field mix" over time and are in part a cause of this change) because men and women typically differ greatly in their field preferences. To begin with the basic proportions, we see in Figure 13 and in Table 4 the changes in the overall proportion of PhD's who are women from 1900 to 1974. In both figure and table, the data are given for 5-year periods, except for the last 5 years, where the explosive growth in proportions of women, year by year, is shown in detail. This proportion, combined with the in- creasing numbers of PhD's during the past quarter-century, results in varying numbers of women, as depicted in "tree ring" format in Figure 14. Here we see the increasing segment attributable to women since 1950, together with the widening rings, as the number of doccorates expands. Looking back toward the center of the graph, we note that there was a rather wide wedge representing women in the 1920"s and 1930's, gradually shrinking in the 1940's, but drasti- cally shrinking during the "GI" period after World War II. SEX DIFFERENCES IN FIELD MIX The differing field mix of men and women doctor- ates is shown graphically in Figure 15, in which the outer ring depicts the total number of doc- torates granted to men since 1920, while the inner ring shows the number of doctorates granted to women. The area of each ring is proportional to the number of doctorates, while the segments within each ring represent the proportions of the several major fields of specialization. Figure 15 also incorporates small tables showing the numbers of male and female PhD's, together with percentages, and also the relative propor- 20 15 o X in O £ o. o K 10 Women as a percentage of all PhD's climbed from 1900 to the late 1920's, then dropped until the mid-1950's, and have risen rapidly in recent years Ql I I I I I I I I I I I I I I 1900 1910 1920 1930 1940 1950 1960 1970 YEAR SOURCE- NRC, Commission on Human Resources FIGURE 13 Women PhD's, 1900-1974. tions of the male and female populations in the several fields. The most obvious sex difference is in the natural science segment. The outer ring (men) is approximately half (50.4 percent) natural sciences, including mathematics and engineering, shown as the shaded portion. The inner ring has only about one-quarter shaded, showing that the natural sciences, mathematics, and engineering TABLE 4 PERCENTAGE OF U.S. PhD's WHO ARE WOMEN, 1900-1974 Women Women Women Period » Percent Period S Percent Period N Percent 1900-1904 150 8.8 1930-1934 1,755 14.7 1960-1964 6,606 10.8 1905-1909 188 9.7 1935-1939 2,026 14.8 1965-1969 13,520 12.4 1910-1914 286 11.3 1940-1944 1,984 13.5 Single Years 1915-1919 324 11.3 1945-1949 2,139 13.4 1970 4,378 13.9 1920-1924 634 15.1 1950-1954 3,617 9.4 1971 4,985 15.0 1925-1929 1,193 15.4 1955-1959 4,647 10.5 1972 5,723 16.6 1973 6,371 19.0 1974 6,785 20.5 SOURCE: NRC, Commission on Human Resources.

18 In "tree ring" terms, women PhD's represent a rapidly widening segment of the circle of recent PhD growth SOURCE: NRC, Commission on Human Resources FIGURE 14 Proportion of women PhD's depicted as tree rings. attract only 24.5 percent of the women. Another prominent sex difference is in education. About one man in six among the doctorate recipients has his degree in education; among the women this proportion is almost doubled (27.9 percent). Languages and literature are smaller segments and hence less conspicuous, but the sex differ- ence is actually larger proportionately: 5.7 percent for the men versus 15.2 percent for the women. In psychology, we find 6.0 percent of the men and 11.2 percent of the women. In the life sciences, the proportions are almost in balance, 16.8 percent of the men and 15.9 per- cent of the women. In the EMP fields,,the dis- parities are greater, ranging from 3.8 percent versus 2.0 percent in mathematics to 10.8 percent versus 0.4 percent in engineering. DOCTORATES GRANTED IN FIELD GROUPS The various fields and field groups have not grown uniformly over time, as has been shown. More detail with respect to the different growth rates, and the consequences in terms of field mix, are explored below. Figure 16 gives an overall picture of the changing output numbers by general field groups. The heavy line shows the growth of the EMP group. The largest single group shown in Figure 16, it also depicts the general growth curve, with a slowing down in the depression and World War II periods, the sharp postwar spurt, the secondary slowing down, then the extended high growth during the 1960's, and, finally, a slower growth during the 1970's— a familiar picture shown in a different form earlier in this chapter. The vertical axis in Figure 16 is average number of degrees granted annually over each 5-year period. Although the other field groups in Figure 16 do not follow exactly the same growth pattern, the major effects of circumstances are similar. The other four fields originally are quite dif- ferent in numbers of doctorates granted, then merge indistinguishably for a period of about 15 years in the 1950's and 1960"s, to emerge later in a different rank order. In 1920 the

19 Outer Ring - M«n Inner Circle - Women I Natural Sciencn 421.072 66.363 487,435 86.6% 13.4 100.0% The "field mix" for men emphasizes the natural sciences; for women, psychology and the nonscience fields predominate I Behavioral Sciences and Nonscience Fields Field Male Female Life Sciences 16.8 16.9 Math 3.8 2.0 Physics 6.2 1.2 Chemistry 10.5 4.6 Earth Sciences 2.3 0.4 Engineering 10.8 0.4 TOTAL. Natural Sciences 50.4 24.5 Field Male Female Education 16.2 27.9 Languages and Literature S.7 15.2 Other Humanities 7.8 9.9 Professions 4.1 3.5 Psychology 6.0 11.2 Social Sciences 9.6 7.8 TOTAL, Social Sciences. Arts, end Education 49.4 75.5 SOURCE: NRC, Commission on Human Resources FIGURE 15 Field mix by sex, 1920-1974.

20 10,000 5.000 rr 1,000 Ul I I 500 100 so Differences in growth rates have shifted rank orders of fields: education, once last, is in a strong second position today EMP Fields Human ities& Professions Life Sciences Behavioral Scie Education 1920 1930 1940 1950 YEAR 1960 1970 FIGURE 16 Growth curves of field groups, 1920-1974, by 5-year periods. rank order of these field groups was EMP, human- ities and professions, life sciences/ behavioral sciences, and education. In 1974 the rank order was EMP, education, humanities and professions, behavioral sciences, and life sciences. The humanities and professions group (here combined to avoid cluttering the graph further) were originally the second largest of the field groups. But this field group underwent a pro- longed period of slow growth and negative growth, to emerge again in recent years below education, which moved up from a poor fifth position to second after the EMP group. Even during World War II education continued to grow, a function of two factors: the large proportion of women in the field and the relatively advanced age at doctorate in the education field, both factors diminishing the effect of the draft. The con- tinued growth of the EMP fields during the World War II period was due to a quite different reason—the vital importance of these fields to the war effort. The life sciences, third in the period from 1920-1950, grew relatively slowly from 1950 to 1974, finally appearing as the smallest of the field groups shown. The be- havioral sciences generally remained one of the smaller field groups until the last 5-year period, when they grew rather rapidly, over- taking the life sciences fields (see Figure 8) . For those interested in the finest detail of subfields. Appendix 1 provides data for the entire 1920-1974 period by fine field, with additional columns for the 1960-1969 period and annual data for the 1970's. CHANGING PROPORTIONS OF FIELD GROUPS The shifting growth patterns depicted above result in varying proportions of the PhD total, as shown in Figure 17, here reduced to four general field groups for the sake of simplicity. The brackets at the sides of the figure show the percent that each of these groups represents in the 1920's and in the 1970's. Although the EMP group has remained relatively constant through most of the half-century depicted here, and actually increased for a time, the recent sharp drop in output has cut the proportion to

21 1001- In proportions of the total, the science/nonscience border has shifted-first upward, then downward, and up again over a halt-century 80 o 0 c. Q. u. o UJ O rr 60 40 -s. O 20 31 — 21 - 145 333 0L 27 EMP Fields 152 Life Sciences 170 Behavioral Sciences 403 All Nonscience Fields Combined I I I I I I 1 I 1920 1930 1940 1950 1960 1970 YEAR SOURCE: NRC, Commission on Human Resources FIGURE 17 Changing proportions of four general field groups. 27.1 percent of the total during the first half of the 1970's, from 31.2 percent in 1920. The life sciences as a group have gradually shrunk from 21.0 percent at the beginning to 15.2 per- cent at the end. The behavioral sciences, which include psychology and the various social sciences, after a quick expansion in the early 1920's, shrank gradually as a proportion of the total, then expanded during World War II and the subsequent period, shrank again during the 1960's, and finally expanded sharply in the most recent period. Nonscience fields show the clearest trend, rising, then falling again until the 1950's, and expanding rapidly in recent years. The overall changes shown in Figure 17 are best understood by examining in more detail the various subfields. In Figure 18, the six fields that compose the natural sciences are shown as proportions of the natural science total. Al- though the changes in the fields at the top and bottom of the graph are most easily visualized, the changes for all fields over the 50-year span are shown by the numbers in brackets at the sides. The proportion attributable to mathe- matics has almost doubled; the proportion within physics has shrunk, then expanded again to about its original size; chemistry has shrunk, except for the period of the 1930's, and now is con- siderably less than half its original proportion (34.3 percent down to 14.6 percent); the earth sciences have diminished gradually from 6.3 percent to 4.1 percent, while engineering has expanded enormously—by a factor of 9, actually— from 2.7 percent in the 1920's to 24.8 percent in the 1970's. Life sciences, as indicated above, have gradually shrunk from 40.3 percent to 35.9 percent, but show some signs of revival in the latest period. The numbers for Figure 18 are found in Table 5. The nonscience fields are shown in Figure 19. At the top, the languages and literature group is shown, with an almost steady decrease in pro- portion of the total of the nonsciences, from the 1930's to present. The other fields within 100 r- 80 - CO HI U UJ K 3 60 UJ o 2 4° a. UJ 3 •s. O 20 11.2 343 In the scientific fields, chemistry has shrunk as a fraction of the total, for more than 40 years, while engineering has burgeoned 14.6 Chemistry 403 1920 1930 1940 1950 YEAR 1960 1970 SOURCE: NRC. Commission on Human Resources FIGURE 18 Changing proportions of fix science fields.

22 TABLE 5 CHANGING PROPORTIONS OF SIX FIELDS IN THE NATURAL SCIENCES AND ENGINEERINGS-YEAR PERIODS, 1920-1974 EMP Field Group Earth Life Chem- Sci- Engi- Sci- Period Math Physics istry ences neering ences Total 1920-1924 5.2 11.2 34.3 6.3 2.7 40.3 100.0 1925-1929 6.4 11.1 32.1 5.2 4.6 40.5 100.0 1930-1934 6.8 10.2 30.2 5.2 7.0 40.5 100.0 1935-1939 5.6 10.9 32.3 4.8 6.2 40.2 100.0 1940-1944 4.8 9.0 36.6 3.3 5.9 40.4 100.0 1945-1949 5.9 10.1 32.5 3.9 12.6 35.0 100.0 1950-1954 5.5 13.0 27.0 3.9 14.4 36.2 100.0 1955-1959 6.0 11.9 24.2 4.6 15.2 38.2 100.0 1960-1964 7.4 12.3 20.4 4.7 21.1 34.1 100.0 1965-1969 8.5 12.4 16.7 4.2 25.4 32.7 100.0 1970-1974 9.0 11.5 14.6 4.1 24.8 35.9 100.0 Percentages may not total 100.0 because of rounding. SOURCE: NRC, Commission on Human Resources. the humanities have also diminished, but not as spectacularly, while the professions, always a small group, have fluctuated somewhat but with- out any marked change in overall proportion. The graph is dominated, however, by the high per- centages in education, a field that has in- creased, with the exception of a single 5-year period, throughout the half-century shown, until it is half of the nonscience total. Table 6 provides the figures. These data on proportions are all brought together and are combined with data on actual numbers of doctorates per 5-year period, in the tree ring graph of Figure 20, in which the field groups are shown as segments of the whole circle. Because the natural science fields are shown on TABLE 6 RELATIVE PROPORTIONS OF TWO GENERAL GROUPS, 1920-1974, 5-YEAR PERIODS: (A) BEHAVIORAL SCIENCES AND (B) HUMANITIES, PROFESSIONS, AND EDUCATION A. Behavioral Sciences B. Humanities, Professions, and Education Language and Literature Other Human- ities Period Psy- chology Social Sciences Profes- sions Educa- tion Total 1920-1924 1925-1929 1930-1934 35.6 33.0 30.8 64.4 67.0 69.2 30.8 24.2 27.9 34.1 31.7 28.0 12.9 13.4 11.7 22.2 30.6 32.4 100.0 100.0 100.0 1935-1939 1940-1944 1945-1949 31.8 29.0 33.1 68.2 71.0 66.9 32.0 26.7 21.7 24.7 22.5 22.4 10.0 10.5 10.7 33.3 40.3 45.2 100.0 100.0 100.0 1950-1954 1955-1959 1960-1964 43.4 46.5 44.8 56.6 53.5 55.2 18.9 18.0 17.4 23.3 22.1 23.0 9.3 10.2 11.6 48.6 49.6 48.1 100.0 100.0 100.0 1965-1969 1970-1974 42.7 41.1 57.3 58.9 18.3 17.0 21.4 19.9 11.3 11.0 49.1 52.1 100.0 100.0 Percentages may not total 100.0 because of rounding. SOURCE: NRC, Commission on Human Resources.

23 100 r- In the nonscience fields, the growth in education has been phenomenal, while the humanities have lost ground proportionately, the professions are relatively stable 80 60 o W U V> I § o a: 40 LLJ U 20 308 341 - 129 222 1920 1930 17.0 199 11 0 521 Visualized as segments of tree rings, the humanities and professions have shrunk and education has expanded as growth rings have widened | 1960 1940 1950 YEAR SOURCE: NRC, Commission on Human Resources FIGURE 19 Changing proportions of nonscience fields. 1970 SOURCE: NRC. Commission on Human Resources FIGURE 20 Changing field mix depicted as tree ring segments.

24 500 i- The U.S. general population has grown slowly and the population of living PhD's has grown rapidly over the past half-century 1920 1930 1940 1950 YEAR 1960 1970 1974 SOURCE: NRC, Commission on Human Resources FIGURE 21 Estimated living U.S. PhD population compared with U.S. general population. either side of the vertical radius, they are most easily visualized as entities. The fluctu- ations shown here are a function both of the growth in total numbers and of the proportions shown in Figure 17. The other fields are shown as less regular segments, but the rapidly in- creasing numbers and proportions in education, for example, are unmistakable. The behavioral science segment has remained roughly constant, while the humanities and professions sector has shrunk. THE DOCTORATE POPULATION What is the size of the living doctorate-1evel population? The first approximation to an an- swer to this question is shown in Figure 21, which shows the size of the total and sex- differentiated living doctorate-level populations in the United States from 1920 to 1974. This figure is based on a computer model1 using graduations and the application of age-specific death rates to the graduation data; emigration and immigration of the doctorate-holding popula- tion has been excluded. The death rates, which are significantly lower than those for the U.S. general population, were taken from actuarial data of the Teachers Insurance and Annuity Association. The assumption that all the gradu- ates from U.S. universities remain in the United States is not true, of course; many go abroad after graduation. But this number is to some extent offset by immigrations; in the model shown here the assumption is made that immigration balances emigration. The precise accuracy of this assumption cannot be tested from data cur- rently available, but it is believed to be good enough so that the conclusions are not materially affected. Figure 21 is semilogarithmic—that is, the vertical scale is logarithmic and the horizontal scale (time) is linear. It is the logarithmic nature of the scale that results in the compres- sion that makes the data for both sexes slightly different from that for men alone. Overall, the proportion of women in the PhD population is about 13.6 percent at present; it has varied from nearly 15 percent in 1940 to less than 12 percent in 1960. The logarithmic scale results in a compression of these numbers by a factor of about 8, when the male and total data are com- 'ihe computer program that produces PhD population estimates begins with data on the distribution of age at completion of the PhD, separately for each sex, field, and time period of graduation—a rather extensive data set. It then calculates survivorship of each age-sex-field group in each year from graduation until all are deceased, using age-specific death rates based on data from Teachers Insurance and Annuity Association. (These rates, quite different from general population age-specific death rates, have been independently verified through application to a known population of scien- tists.) The program then accumulates data across cohorts to provide a table, by age, of the living PhD's of a given field and sex, in any given year. Data are provided for each of 10 fields of PhD and may be accumulated in field and sex groupings as desired. Projections to future years are pos- sible, based on assumed PhD graduation rates.

26 100 50 < 10 CL O Most science fields and the humanities and professions have grown at about 7% per annum, on the average, in number of living PhD's 1 1940 1945 1950 1955 1960 YEAR SOURCE: NRC, Commission on Human Resources 1965 1970 1974 FIGURE 22 Estimated living U.S. PhD population in seven fields compared with U.S. population age 25 and over. pared. Table 7 presents the total data by sex, and Table 8 presents the data by field of doc- torate but with reference data on the general U.S. population. In all of these population data, field of doctorate, rather than field of present specialization, is presented. Switching of fields after the doctorate is not taken into account in these figures. Field switching has been described in a separate report,2 as far as scientists and engineers are concerned, and will be discussed further in Chapter 2. In Figure 21, the growth of the total U.S. population is shown for comparison with the growth in the PhD population. The scale for the U.S. population is shown in the right margin; it uses the same scale as the PhD population scale on the left but is multiplied by 10,000. Over the period from 1920 to 1974, the U.S. 2Cosinission on Human Resources, NRC, Field Mobility of Doc- toral Scientists and Engineers (Washington, D.C.: HAS, 1976). population approximately doubled, going from about 105 million to over 210 million. But over the same period, the PhD population in- creased by a factor of 50, going from 8,830 to 448,900. In terms of proportion, the PhD's in- creased from less than 1 per 10,000 of the gen- eral population in 1920 to about 21 per 10,000 in 1974. Figure 22 depicts the growth of 7 of the 10 doctoral field populations, over the period 1940-1974. In this set of fields, the growth is rather regular, and the curves run approxi- mately parallel. There are differences in growth rate, ranging from an average annual increment of 5.0 percent in chemistry to 7.5 percent in mathematics. As expected on the basis of doctoral graduations, the growth has been steepest over the past 15 yee.rs and, for most fields, slowest during the World War II period. The smallest of the fields shown in Figure 22, earth sciences, increased from about 1,300 in 1950 to about 9,000 in 1974, averaging

26 TABLE 7 ESTIMATED POPULATION OF LIVING U.S. PhD's, BY SEX, 1920-1974, COMPARED WITH U.S. POPULATION Year PhD's per Million of Estimate Male PhD's Female PhD's Total PhD's U.S. Population 1920 7,580 1,250 8,830 106,466,000 83 1925 11,550 1,950 13,500 1930 18,630 3,150 21,780 123,188,000 177 1935 28,900 4,900 33,800 1940 40,700 6,920 47,620 132,122,000 360 1945 51,000 8,690 59,690 1950 67,950 10,930 78,880 151,683,000 520 1955 103,000 14,530 117,530 1960 140,300 19,000 159,300 179,323,000 888 1965 196,800 25,800 222,600 1970 297,700 41,000 338,700 203,200,000 1,667 1974 388,400 60,500 488,900 213,000,000 2,108 (estimate) SOURCE: NRC, Commission on Human Resources. TABLE 8 ESTIMATED PhD POPULATION, BY FIELD, 1940-1974, COMPARED WITH U.S. POPULATION 25 AND OVER Reference Year PhD Field 1940 1945 1950 1955 1960 1965 1970 1974 Mathematics Physics Chemistry 1,460 2,600 7,900 1,630 2,200 4,200 13,380 3,260 6,650 18,190 4,480 9,010 22,880 7,020 12,960 28,750 11,940 19,900 37,580 16,190 25,160 43,640 3,110 10,260 Earth sciences Engineering Life sciences 1,260 1,230 9,580 1,440 1,630 1,800 2,990 15,340 2,520 5,870 22,380 3,450 9,140 4,880 16,720 40,260 7,080 31,450 8,970 12,040 29,870 58,570 43,260 75,200 Psychology Social sciences Humanities and professions Education 2,140 4,710 2,560 5,710 3,520 7,500 6,530 11,090 10,050 14,990 14,580 20,410 22,340 30,390 42,000 30,650 11,770 5,190 14,370 6,940 17,880 24,390 16,660 31,660 23,800 42,670 34,380 63,550 55,660 84,870 79,240 10,080 TOTAL 47,620 59,690 78,880 117,530 159,300 222,600 338,700 448,900 U.S. population 74,775 86,484 99,465 109,899 113,000 Age 25 and over (in thousands) The data have been rounded, and hence may not add exactly to the totals given. SOURCE: NRC, Commission on Human Resources. a growth rate of 6.3 percent per year. The largest field, humanities and professions, had almost 12,000 in 1940 and grew to over 84,000 in 1974, averaging a growth rate of 5.9 percent annually. The growth rates for the other fields, over the period shown, averaged 7.0 percent for physics, 6.3 percent for the life sciences, and 6.5 percent for the social sciences. In Figure 22, the PhD population by field is compared with the U.S. population age 25 and

27 100 60 I I c O 10 a. 2 a Education and psychology, growing at 8% per year, have exceeded moil other fields; engineering has averaged about 11 % per year 1 1940 1945 1950 1955 1960 1965 1970 1974 YEAR SOURCE: NRC. Commission on Human Resources FIGURE 23 Three fast-growing PhD populations. over. This is a more relevant reference group than the total U.S. population shown in Figure 21, since almost all PhD's are over the age of 25. Again, the general population scale is at the right, and again it is multiplied by a factor of 10,000 as compared with the PhD's. About the same relative difference is apparent in the slopes of the PhD populations, as compared with the U.S. 25-and-over totals. However, the per- centage differences vary. In 1940, there were about 6 PhD's per 10,000 of the population 25 and over; in 1950 this ratio increased to slightly over 9; in 1960, to 16; in 1970, to al- most 31; and in 1974 the ratio was almost 40 per 10,000 U.S. population of comparable age. Since'slightly more than half of the general population over 25 is female, while about 86 per- cent of the doctorate population is male and 14 percent female, the PhD/population ratio for males is about 70 per 10,000; for females about 10 per 10,000. Figure 23 depicts the growth of the remaining three fields of doctorates. These are all faster growing than those shown in Figure 22 and, if superimposed, would cross the lines of that figure repeatedly. The three fields are educa- tion, psychology, and engineering. Education, with an average annual growth rate of 8.4 per- cent, grew from about 5,140 in 1940 to about 78,800 in 1974. Psychology, with an average growth rate of 8.2 percent, rose from about 2,200 to 30,300 over the 34-year period. Engi- neering, with a growth rate averaging 11 percent per annum, moved from the position of smallest field (about 1,260) in 1940 to one of the largest (43,200) in 1974. As in Figure 22, the total U.S. population age 25 and over is shown for comparison. The detailed data, showing the numbers in each field by sex and by single years of age, for each year from 1920 through 1974, are avail- able in computer tape form and are the basis for additional analyses described in Chapter 2 re- lating to demographic data.