College and Research Libraries M. S U S E E L A S U B B A R A O The Place of Library Resources in Doctoral Programs This paper supplements an earlier paper by Robert B. Downs on doctoral programs and library resources. While the factors involved in successfully implementing a doctoral program are many and complex, to carry out such a program in a variety of fields, it appears that there should be at least three thousand current periodicals (and five hun- dred thousand volumes, as Dr. Downs states). Even with the best library resources, one cannot generally hope to produce more than one doctorate out of every ten graduate students enrolled in any year, as figures in this paper indicate. I N A RECENT P A P E R entitled "Doctoral Programs and Library Resources," R. B. Downs 1 presented a table showing the number of doctoral degrees—minimum being five—conferred by all the educa- tional institutions in the United States during the decade 1953-1962. The pur- pose of that paper was to consider if di- rect correlation exists between the num- ber and variety of doctoral degrees awarded and the strength of library re- sources in individual institutions. One will easily agree with Dr. Downs that "since there are no established norms, exactly how many volumes should be held by the library and how much mon- ey spent for books in an institution offer- ing doctoral programs are debatable matters. Pragmatically speaking, how- ever, it seems doubtful that high level doctoral work in a variety of fields can be carried on with less than half a mil- lion volumes." The purpose of the present paper is to consider briefly some of the other im- 1 "Doctoral Programs and Library Resources," CRL, XXVII (March 1 9 6 6 ) , 1 2 3 - 2 9 . Mrs. Subbarao is Catalog Librarian in the University of Alberta. portant factors upon which the produc- tion of doctoral degrees in an institution depends and the relative importance of library resources—books and current pe- riodicals—as one of these factors. (Dr. Downs did not take into account the strength of current periodicals holdings as a factor, and his table does not give figures on this.) We also consider the proportion of doctorates conferred to the total number of graduate students en- rolled. It is a simple but important fact that the number of doctorates produced by an institution, or by one of its depart- ments, depends almost entirely upon the number of graduate students who work for the doctorate degree. Also, if the concerned institution offers a doctoral program, the size of the doctoral stu- dent body has some, but by no means vital, relation to the size of the graduate student body as a whole. The number of admissions to the graduate school of an institution for a master's program de- pends upon various factors. As some of these, we might mention organization and admission policies, curricula, physi- cal facilities, size of the undergraduate student body in the concerned institu- tion and the other institutions in the 4 2 4 / Place of Library Resources in Doctoral Programs / 425 neighborhood, the amount of better em- ployment opportunities available in the surrounding community for higher qual- ified persons, financial aid available to graduate-level students, library resour- ces, the size and stature of its facul- ty members, and the general reputation and standing of the institution in the academic world. Some of the above- mentioned factors of course may have little or no relevance or importance in a specific institution. However, it would seem reasonable to assert that the last five of these factors have a direct bear- ing on the number of doctorates pro- duced. An institution may, as a part of its policy, emphasize its interest in and duty toward the advancement of knowl- edge and promotion of research. In pur- surance of this policy, it may rapidly provide excellent physical facilities and even library resources. But it may still be unable to attract, even with the best of efforts, the right type of faculty mem- bers. This is the one thing that cannot be achieved as a crash program. But if it is successful over the years in securing a distinguished and widely recognized cadre of faculty members, the institution is indeed lucky, and it can be said to have crossed the main hurdle. Faculty members who devote a good deal of their time to research and writing will surely stimulate their graduate students with new and recent ideas. The pres- ence of such a faculty in a department is doubtless the most important factor in determining the size of its doctoral students and the number of doctorates produced. Graduate students flock to such a department, for they know that they will be benefited and inspired by the presence of such a staff. From di- rect exchange of ideas they get the needed stimulus for creative scholarship, and at least some of the excitement of research going on in the department "rubs off" on the receptive minds of the students. If the kind of faculty described above is available in a department which is offering a doctoral program, then the size of the library collection, even in the concerned departmental library, has really little relevance to the number of doctorates produced by the department. A good staff will surely see that the li- brary collection pertaining to their field and to their discipline is adequate, and their doctorate production is bound to be good—not because of their library resources, but because of the quality of the staff (and students). A not-so-good staff in a department provided with suf- ficient financial resources can build up a vast library collection, but the produc- tion of doctorates may not be high. The existence of good library facilities is gen- erally, at best, a necessary condition for the production of doctorates, but not at all a sufficient condition. In exceptional circumstances, this may not even be a necessary condition. This is because a thesis adviser may suggest research problems for which there is not much existing literature, or the student may gather all existing literature on a par- ticular topic of research by means of re- prints and preprints from the concerned authors, and then go on with his re- search with no more trips to the library! But, on the whole, one agrees with Dr. Downs that "an institution outstanding for its graduate offerings is almost invar- iably equally notable for the strength of its library resources." The converse, of course, is not true—and this could be for various reasons. A simple reason could be that the institution does not offer a graduate program at all, even though it might have outstanding faculty mem- bers—like the Institute for Advanced Study at Princeton. A second reason may be the lack of sufficient research ori- ented faculty members. W e might mention here a rather pe- culiar situation presented by some of the state colleges in the United States. Some have very decent library resources and 426 / College 6- Research Libraries • September 1968 a good number of research-minded fac- ulty, but offer nothing beyond master's programs. Fortunately, such faculty put these resources to good use in connec- tion with their own research work—even though they are not required to engage in research as a requisite for advance in salary or rank. The table given in Dr. Downs's paper indeed provides some very curious facts. Harvard holds the first rank in the num- ber of library volumes, but only fifth rank in the number of doctorates con- ferred, while the corresponding ranks for Columbia are sixth and first. (All these remarks pertain only to the decade 1953-1962, and the position could pos- sibly have changed since then.) Wiscon- Fig. 1.—Correlation between number of doctorates conferred and number of volumes in in- stitution's library, 1953-62. Place of Library Resources in Doctoral Programs / 427 T A B L E 1 I N S T I T U T I O N D O C T O R A T E S C O N F E R R E D I N 1 9 6 2 - 1 9 6 3 G R A D U A T E S T U D E N T S E N R O L L E D 1 9 6 2 - 6 3 C U R R E N T P E R I O D I C A L S I N S T I T U T I O N Rank of Institution in This Respect No. Per Cent of All Grad. Students Rank of Institution in This Respect No. C U R R E N T P E R I O D I C A L S California . 1 731 4 1 18,358 80,000 (all campuses) Columbia . . . 2 517 5 4 9,854 not 9,854 available Illinois 3 450 6 9 7,115 15,903 Harvard 4 446 6 8 7,617 25,883 Wisconsin . 5 428 9 13 4,862 9,076 Michigan 6 370 4 3 10,412 21,290 N.Y.U 7 345 2 2 18,620 5,000 Ohio State . . . 8 329 8 15 4,373 16,200° Minnesota . 9 314 6 12 5,647 12,432 Purdue . . . . 10 290 7 18 3,871 11,682 Stanford 11 276 6 16 4,340 26,096 M.I.T 12 274 9 20 3,142 3,215 Indiana . . . 13 266 3 5 8,969 12,000 Yale 14 231 6 19 3,772 6,000- Chicago . . . . 15 227 5 14 4,779 55,000 Texas 16 223 9 23 2,590 6,427 Mich. State 17 223 3 10 6,587 13,173 Cornell . . . 18 206 7 21 3,102 13,361 Iowa 19 205 8 22 2,721 6,750 Penn. State 20 202 8 24 2,549 8,000 Iowa State 21 174 11 26 1,662 9,470 Princeton . . . . 22 171 18 27 956 16,000 Northwestern 23 168 7 25 2,419 19,327t Pennsylvania 24 164 2 7 7,966 8,500t Univ. of Wash. 25 161 4 17 4,317 17,300t S. Calif 26 147 2 6 8,178 6,904 Pittsburgh 27 138 2 11 5,661 9,534 ° Serials omitted, t Some serials included, t Some newspapers included. sin holds third rank in the number of doctorates conferred, but only sixteenth in the number of library volumes held. Duke holds more library volumes than Wisconsin but takes only the thirty-sev- enth rank in the number of doctorates conferred. For every one hundred li- brary volumes held by Purdue, Miami holds more than one hundred and nine volumes; but for each doctorate pro- duced by Miami, Purdue produced 203 doctorates. Thus Purdue gets the thir- teenth rank in this respect to Miami's 169th! If there is a perfect correlation between the ranks by number of doc- torates conferred and number of library volumes held, an institution holding a certain rank with respect to the former should hold the same rank with respect to the latter. Plotting the points corre- sponding to each of these institutions with these ranks taken, respectively, as the x and y coordinates, the resulting graph would, under a perfect correla- tion, be a straight line through the origin equally inclined to the x and y axes. The actual graph for the institutions which produced a minimum of a thousand doc- torates during the decade under con- sideration is shown in Figure 1. The reader can see for himself how widely this graph differs from a perfect correla- tion, y = x. What is the other available data then that may have relevance to the number of doctorates conferred by an institu- 428 / College 6- Research Libraries • September 1968 Fig. 2.—Correlation between number of doctorates conferred and number of graduate stu- dents enrolled, 1962-63. tion? Since doctoral programs invariably involve writing a thesis, there should be, of course, facilities in the library for re- search, and this is certainly something indispensable. Since journals usually provide a greater source of current re- search material than do books, it is rea- sonable to consider if there is some cor- relation between the number of doctor- ates conferred and the number of cur- rent periodicals it receives. Another fac- tor which may have a bearing, at least in the case of the big institutions, is the to- tal number of graduate students enrolled in any year (both for master's and doc- torates) in relation to the percentage of doctorates awarded. Table 1 gives: ( 1 ) the total number of Place of Library Resources in Doctoral Programs / 429 graduate students enrolled in the Fall of 1962; ( 2 ) the number of doctorates con- ferred during the period September 1962-June 1963; ( 3 ) the percentage these doctorates form out of the total number of these graduate students; and ( 4 ) the number of current periodicals in the concerned institution. This table is confined to the twenty-seven institutions which conferred at least a thousand doc- torates during the decade 1953-1962. The table was prepared to determine if these data are helpful in drawing any significant conclusions, at least in the case of these leading institutions. Using the table, a graph (Fig. 2) was prepared that correlates the rank of the institution by virtue of number of doc- torates conferred (plotted on the x co- ordinate ) with its rank by virtue of num- ber of graduate students enrolled dur- ing 1962-63 (plotted on the y coordi- nate). A similar graph using the total number of periodicals would have been useful; however, this could not be done since the figures available (as given in American Universities and Colleges, 1964 edition) sometimes include serials —either wholly or partly—and some- times do not. Table 1, like Dr. Downs's, shows some curious facts and gives some useful in- formation. Except in very few cases, the number of doctorates conferred by an institution during 1962-1963 significant- ly exceeds the average number con- ferred for the same for the decade 1953- 1962. This is doubtless to be expected with growing enrollments in the gradu- ate schools, with larger numbers of re- search-minded faculty being appointed to the institutions, and with a growing emphasis on the importance of research degrees. Also, the first eight institutions in Dr. Downs's table are still the first eight in our table, though with small changes in their relative order. Wiscon- sin has fewer periodicals and fewer graduate students than Pittsburgh but has awarded more than three times the T A B L E 2 Doctorates con- ferred, 1962- 1963, as per cent 18 11 9 8 7 6 5 4 3 2 of all graduate students . No. of institutions 1 1 3 3 3 5 2 3 2 4 number of PhD's and occupies in this re- spect the fifth rank to Pittsburgh's twenty-seventh. None of the institutions has fewer than three thousand current periodicals. Turning to the percentage of graduate students who got doctoral degrees dur- ing the year under consideration, Prince- ton has the highest figure, 18 per cent, while having the lowest number of grad- uate students. The next highest is 11 per cent for Iowa State, and curiously this institution's graduate student enrollment is the second lowest. The lowest per- centage is 2, while the percentage held by the largest number of institutions is 6. These details are summarized in Ta- ble 2. There may be many reasons for Princeton having the smallest enrollment of graduate students and yet producing the highest percentage of PhD's. As their graduate school Announcement says, their admissions are normally limited to male students, and the number of stu- dents in the graduate schools is strictly limited. With its excellent faculty and high reputation in the academic world, it will receive a large number of appli- cants, but, as their Announcement says, they choose the most outstanding among the applicants. There is no program at Princeton designed for students who wish to take the degree of Master of Arts as a terminal degree. The master's degree is granted there as an incidental degree, and is offered after completion of a portion of the requirements of the PhD degree. There are some other insti- tutions which adopt roughly the same admission policies as Princeton, though the percentage of PhD's produced is not 430 / College 6- Research Libraries • September 1968 as high as that for Princeton. As an in- stance, we might mention Yale. Its Bul- letin says that the size of each depart- ment is strictly controlled, and that, ex- cept for programs in industrial admin- istration, international studies, and teaching, it gives preference in admis- sion to candidates who intend to com- plete the PhD degree. It is of interest to notice that our table shows that Yale's percentage of PhD's out of a total grad- uate student body of 3,772 is only 6— which is a third of Princeton's. And Yale's faculty and academic standing are generally considered as good as any- one's. There may be a number of other factors involved which need to be con- sidered. However, one can still draw some useful conclusions. Keeping aside Princeton and Iowa State as exceptions, the ratio of the number of PhD's con- ferred to the total number of graduate students enrolled is in all cases less than 1:10, and in more than half the cases it is even less than 1:16. This being the po- sition in the top twenty-seven institu- tions, one can safely assume that things are no better in the cases of the other in- stitutions. In summary, w e can say the follow- ing. To provide for an effective doctoral program and to hope to produce a de- cent number of PhD's annually, it would appear essential for an institution to have a book collection of at least half a million volumes (as Dr. Downs con- cluded); a periodicals collection (cur- rent subscriptions) of at least three thousand; and admissions policies which allow a graduate student body which is at least ten times the number of PhD's it wishes to produce. But these, among many others, are only strictly necessary factors for successful implementation of doctoral programs. After a certain stage is attained, the number of library vol- umes held or the number of current pe- riodicals becomes less and less signifi- cant as a factor in the number of doc- torates produced. ( T h e amount of money spent by the institution on improving its library resources is reflected by the num- ber of books and current periodicals held, and it need not therefore be con- sidered as a separate factor.) Again, merely trying to multiply the number of graduate students does not increase the output of PhD degrees. What matters most, after the above necessary condi- tions are met, are ( 1 ) the number of scholarly and research oriented faculty members w h o are active in publication and capable of inspiring and guiding the graduate students for doctoral work, and ( 2 ) the importance the institution's ad- ministrators assign to securing, retaining, and aiding such faculty. This of course depends in turn upon the financial re- sources of the institution, the availability of such qualified persons for recruitment, and various other factors which are be- yond the scope of the present paper. If we had data (qualitative as well as quantitative) from each institution on the number of faculty members who are active in research, it would doubtless have provided a very significant factor in relation to the number of doctorates produced annually by the institution. One could only wish that such data were readily available.