2000-2001 Report of the Committee on Placement
Division of Professional Matters of the American Philological Association
Please note that the supporting documentation for this report appears in linked tables that will appear in resizable pop-up windows so that they may be consulted alongside this report; these pop-up windows should be closed as you finish consulting each table. The full report with tables at the end is available as an Adobe Acrobat file that may be downloaded and printed (printable version); it is best to use the “Print as Image” command when printing Acrobat files.
Part I. The Status of the Job Market
We are happy to report that the improvement in the job market we noted in our last report (October 2000 APA Newsletter, Adobe Acrobat file) has continued into the 2000-1 placement year. Preliminary figures for Fall 2001 suggest that this trend is also continuing into the current placement year, even despite the downturn in the general economy.
The total number of candidates registered with the Placement Service for 2000-1 was 415 (58% male, 42% female), compared to 443 the previous year and an all-time high of 596 in 1995-96. However, a higher number of these (297 or 72%) attended the annual meeting in San Diego, compared to 273 or 62% the previous year. This could indicate that more of those who do register with the Placement Service are committed job seekers, whereas some of those who already have jobs and in the past registered merely to browse the ads or apply very selectively for better jobs are now relying on the APA website instead of registering.
At the other end of the process, the number of positions advertised with the Placement Service continues to increase. In 2000-1 the Placement Service received 196 positions postings (176 definite, 20 possible), compared to 186 the previous year (169 definite, 17 possible). The formula that has been used in the past to compare numbers of candidates to numbers of positions adds 2/3 of the possible positions to all the definite positions, while counting all candidates registered with the Placement Service; this yields a ratio of 2.19 candidates per position for 2000-1, compared to 2.45 in 1999-2000. This number has steadily decreased from the all-time high of 4.55 in 1994-95. It is in fact the lowest ratio since our records begin; see Table 1 and Figure 1 for historical perspective. [Note: the vacancies number has been decreased slightly from the initial publication of this report (from 194 to 189) because 5 unadvertised positions were inadvertently included. All ratios and other calculations using the vacancies figure have also been recalculated in both the text of the report and the tables.]
Year-to-date data for 2001-2 suggests an acceleration of this trend: 337 candidates had registered with the Placement Service as of November 27, 2001 (compared to 366 as of the same date a year earlier), and 137 positions have been advertised (compared to 111 as of the same date a year ago).
It should be emphasized that there are still far more candidates than jobs. Moreover, even positions advertised as "definite" are sometimes not filled in the same year they are advertised, either due to protracted searches or administrative freezes: for instance, of the 166 definite positions advertised in 1998-99, at least ten were not filled. However, there are also many positions, particularly of a part-time or last-minute nature, that are never advertised through the Placement Service. Correspondingly, some candidates do not register with the Placement Service, particularly if they are not APA/AIA members of if they apply for a small number of positions. Of the 174 named hires announced to the APA, all but 5 (169) were among the 196 positions advertised with the Placement Service; the anonymous survey revealed an additional 35 hires for candidates, bringing the total of 2000-1 hires known to the APA to 209. However, 66 (34%) of these 196 positions, however, went to individuals who were not registered with the Service. This helps to explain why only 34% of all Placement Service candidates, and 46% of all Placement Service candidates who attended the annual meeting, obtained 2001-2002 positions (at least according to the information available to the APA, see Table 16). It would therefore be mistaken to treat these figures as definitive or absolute reflections of the actual market, but the trend lines they reveal are nevertheless real.
All things considered, the 2.19 ratio of candidates to positions (recorded for 2000-1) probably overstates the actual labor supply: many who register with the Placement Service do so merely to browse the ads and apply for few if any positions. A more accurate ratio might be calculated based on candidates attending the annual meeting relative to positions: by this criterion, the 2000-1 ratio is only 1.57. Even this figure includes some graduate students who may have put themselves on the market prematurely (called "testing the waters") and many candidates who already had positions, but were merely applying for better positions. According to Table 5A, over 43% even of those who attended the convention applied for ten positions or fewer, suggesting that they may already have continuing positions and were in the market only selectively. If we subtract this 43% from the candidates at the annual meeting, the ratio of the most earnest candidates to positions available drops to 0.9, i.e. fewer candidates than jobs.
Another way of assessing the balance of labor supply and labor needs is to compare the number of permanent positions advertised to the number of new Ph.D.s produced in a given year. During 2000-1, 82 permanent (i.e. tenured or tenure-track) positions came open, while only 61 new dissertations were completed, at least to judge by those listed in the June, August, and October 2001 APA Newsletter. However, we find that some institutions are careless or dilatory in reporting dissertation information to the APA, so the actual number of dissertations completed is probably somewhat higher. Another way of calculating new Ph.D.s in a given year is to examine the number of Placement Service registrants who attend the annual meeting and list their degree as expected during the following year: this number was 73 during 2000-1. By both measures, we see that the number of permanent positions in the field is greater than the number of new classicists being produced by U.S. and Canadian doctoral programs.
We therefore see confirmed in the figures for 2000-1 a situation that we predicted in our last report, namely the development of a serious labor shortfall: i.e. not having enough qualified candidates to fill the positions available. The long-predicted wave of faculty retirements on the part of those who entered the profession during the boom years of the 1960s is finally beginning, at the same time that demographic trends are expanding student populations in many public universities. These developments also come at a point when many major Ph.D.-granting departments are experiencing a significant decline in the number and quality of graduate school applications, as well as an increase in attrition among advanced graduate students who elect to leave the program and participate in the economy in other ways. We therefore regard it as imperative that all classicists, even those teaching in strictly undergraduate programs, should disseminate the news that there is no longer an unemployment crisis in our field, but that opportunities in higher education are likely to be abundant in the coming years. We should all redouble our efforts to encourage talented undergraduates to consider graduate school and teaching as highly practical career choices at the present time. Show them this report!
It is equally important, however, to impress the significance of these trends on high administrators within our institutions. To some extent, the same trends are apparent in all academic disciplines, but Classics may be in a more serious position of labor shortfall than some other fields. Administrators must be told that they can no longer rely on a saturated job market to guarantee an adequate supply of energetic and creative personnel to staff the positions that will be coming open over the next decade. If we are to recruit the talent we need into graduate school and keep them there, we need better graduate student support and, perhaps even more importantly, faculty salaries at the end of the road that will be more competitive with what those students could realize in the computer industry and other learned professions, many of which they can enter with fewer years of training than a Ph.D. demands. It will be news to none of us, particularly those who teach at public universities, that our salaries have not, on average, even kept pace with inflation over the last 20 years. It will also be news to few of us that classicists tend to be the "lowest of the low" in terms of faculty salaries even within Liberal Arts. We continue to believe that the APA should make it a major priority to study faculty salaries within our field, relative to other disciplines, and explain to university administrators the serious impact that chronically uncompetitive salaries have on the profession’s ability to recruit sufficient talent.
Two notes of caution should be added here. It is difficult to project trends several years in advance. Although the present economic downturn does not seem to have affected the employment market in our field, the relatively mild recession of the early 1990s had a very severe negative impact, as is clear from the historical data in Table 1. And although the number of positions advertised increases each year, much of that improvement has come in the form of short-term or part-time positions. In 2000-1, 54.4% of the positions advertised with the Placement Service were non-permanent (i.e. not tenured or tenure-track), whereas only 49.6% were in 1999-2000. The increased competitiveness of the labor market for hiring institutions has not yet translated into an improvement in the quality of positions any more than it has in the level of compensation.
Part II. The Demographics of the Market
Starting in Fall 1999, the Committee on Placement began to collect detailed demographic information on the candidates registering with the Placement Service, in an effort to determine which factors, if any, were most likely to predict success on the academic job market. In the first year, this took the form of a voluntary questionnaire candidates were asked to return at the same time they submitted scheduling forms for the annual meeting; this data was reported in the October 2000 APA Newsletter. In 2000-1 the same questions were incorporated into the initial registration forms for all users of the Placement Service, although candidates were free not to answer any questions with which they were uncomfortable. The information was entered into an anonymous database and was then coordinated by numerical code with responses to a voluntary survey sent at the end of the year, asking candidates about the results of their job search. This information can also be coordinated with data the APA compiles about job interviews scheduled through the Placement Service and positions filled, as announced by institutions. The result is a much more accurate profile of how each demographic subgroup fares both in terms of interview results and final hires. Last year, we were only able to track the former.
Due to changes in the format and presentation of the year-end survey, we had a better return rate than in past years: 46% of the candidates attending the annual meeting, 40% of the candidates overall. Although female and minority candidates tend to return the survey at a somewhat higher rate than males (54% of females at the annual meeting, 41% of males; 71% of all minorities, 41% of all whites) and single candidates at a somewhat lower rate than those who are married or in long-term relationships (34% of all single candidates, 44% of all married candidates, 69% of those in other types of long-term relationships), we nevertheless believe that the year-end survey yields useful information about those within each category, even if not providing a completely representative overall sample.
As we noted last year, our survey reveals that there is not currently a crisis of vast unemployment in our field (see Table 5A). Of the 415 registrants with the Placement Service, only 23 (about 5.6%) listed their current employment situation as "Non-Academic" or "Unemployed," and only six of those 23 said that they expected to apply for more than ten positions, suggesting that they are either geographically limited or for some other reason very particular about the kind of position they will take. What does exist in our field is a certain amount of "underemployment," in that 43 respondents (about 10.4%) listed their current position as "Part-Time/Adjunct." But even in this group, less than half (17 out of the 43) expected to apply for more than ten positions, suggesting that geographical or other factors played a limiting role in their availability for positions. Most candidates were either current graduate students (111) or in full-time non-tenure track positions (143). Only three of the 30 candidates in tenure-track positions reported applying for more than ten jobs, suggesting that most were trying to find better positions, rather than having to find new positions after being denied tenure. These results confirm the trend lines discussed in Part I of our report: the excess labor capacity to which the profession has been accustomed for many years is rapidly dwindling.
Some of the most useful results of our survey pertain to the factors which are most likely to lead to a higher number of interviews and final hires. Factors which have a major positive impact include being young, being female, being a U.S. citizen or resident, having a completed doctorate from a large and well-recognized American graduate program, and working in certain fields of specialization (especially literary studies). Factors which appear to have less impact include marital status, race, date of Ph.D., and amount of publication.
Let us start with area of specialization, since it is here that we find some of the most dramatic differences (see Tables 6A-B). The 297 candidates at the annual meeting had a total of 678 interviews, for an average rate of 2.3 interviews per candidate. Higher than average rates were observed in Latin literature (3.5), Greek literature (2.8), and comparative literature (2.8). These figures are consistent with what has been observed in previous placement years (see Table 6 in last year’s report and Table 7 in the Committee's report in the June 1999 APA Newsletter). Judging by interview rates, the market for ancient historians and linguists was quite a bit less good in 2000-1 than in 1999-2000 (1.4 vs. 3.0 interviews for Greek historians, 1.6 vs. 2.0 for Roman historians, 0.7 vs. 2.0 for linguistics), but the market for archaeologists (1.4 vs. 0.9 interviews) was somewhat better. The market in ancient philosophy was about the same (1.0 vs. 1.2 interviews). The prospects for candidates in ancient history, philosophy, linguistics, and art/archaeology may not be quite as grim as these statistics suggest, in that some of the jobs in these fields are available in History, Philosophy, Linguistics, or Art History departments, for which the APA/AIA meeting may not be the preferred venue. In relatively small fields like these, the number of jobs available can vary greatly from year to year.
On the other hand, the actual hiring results tell a somewhat different story from the interview results. Of the 297 candidates at the annual meeting, the APA received information that 143 of them were hired into new positions, based either on the candidates’ own testimony in the end-of-year questionnaires or institutional hiring announcements. This yields a rate of 48% new hires, which probably understates the actual rate, since only 46% of the candidates at the meeting returned questionnaires and many institutions either never announced their hire to the APA or never advertised with the APA in the first place. Nevertheless, Table 6B gives the information available based on the hires we know about. Although candidates in Latin literature received the highest number of interviews, their rate of hire was only average. Candidates in Greek literature, comparative literature, and ancient philosophy were hired at above average rates. And although the interview rates for candidates in ancient history and art/archaeology were well below average, their hiring rates were close to average. However, the number of permanent hires (tenured or tenure-track) was definitely better in literary studies (both Greek and Roman) and Roman history. These trends confirm the data in a longer-term study of hiring results compiled by committee member Alexander MacGregor (Ten Years of Classicists: Dissertations and Outcomes , 1988-1997 [Wauconda: Bolchazy-Carducci, 1998]).
Claims that age is a factor in hiring do find some support from our figures (see Tables 7A-B). Here it seems appropriate to eliminate from consideration the candidates who reported applying for ten or fewer positions, since many in that group will be candidates who already have established positions and may thus, on average, be older and have fewer interviews, since they are applying for fewer positions to begin with. To be sure that we are comparing candidates who have applied for comparable numbers of positions, we have compiled figures both for those candidates who applied for more than 20 positions and those who applied for more than ten. In both cases, one sees a clear progression, with an above average number of interviews being granted to those under the age of 30 and a significantly below average number being granted to those over 40. Table 7B shows the same tendency in regard to final hires, but here the differences are somewhat less sharp. It is unclear whether the advantage of younger candidates is due to age discrimination or a preference for candidates who finish the Ph.D. quickly rather than after extended careers in graduate school.
Closely related to the issue of age is the age of one's Ph.D. (see Tables 8A-B). This factor has been studied previously (see Table 8 in the Committee's report in the October 2000 and June 1999 APA Newsletter); for the same reasons as above, more reliable results can be obtained by screening out candidates who have low interview rates because they applied for relatively few positions. While previous years’ results had suggested that candidates whose doctorates were more than five years old were at a disadvantage, that is not clear from the more complete results available this year. Both candidates whose doctorate is not yet complete and those whose doctorate is 5-6 years old do as well as or better than average; while those whose Ph.D. was earned in 1994 or earlier do seem to obtain fewer interviews, the number of candidates in that category is too small to carry much statistical significance. Moreover, Table 8B shows that those who earned doctorates from 1992-94 did at least as well as other candidates in terms of final hires, and actually somewhat better than average in obtaining permanent positions. It would therefore seem that with the shortfall in labor supply, age of Ph.D. is not as much of a factor as it used to be. Moreover, Table 5B suggests that one’s current employment status does not necessarily affect one’s success in obtaining jobs, especially when one considers that some groups tend to apply for fewer jobs.
The institution from which one has earned the Ph.D. does seem to play a significant role in one's ability to attract interviews. Tables 9A-B break down interview results by institution, listing only those Ph.D.-granting programs that had two or more candidates who were applying for more than ten positions. Tables 9C-D give results in terms of final hires. Results for individual institutions should be used with great caution: most institutions had five or fewer candidates, and it is difficult to draw firm statistical conclusions from such small samples, since the quality of candidates an institution produces may vary from year to year. However, if these figures are accumulated over several years, they could provide a useful objective index of how well various graduate programs are thought to train future teachers and scholars. In general, larger and better known graduate programs tend to obtain more interviews for their degree holders than smaller programs, but there are some notable exceptions. Ph.D.s from foreign universities do significantly less well on the American market.
Indeed, citizenship and residency status appear to play a major role in obtaining interviews (see Tables 10A-B). It should be noted, however, that being a foreign national is not a disadvantage as long as one is either a current U.S. resident or has earned one's doctorate from an American university. Otherwise, foreigners find the U.S. job market very difficult, and this disadvantage applies to Canadians not resident in the U.S. just as much as to Europeans. The surprisingly widespread notion that American universities prefer Europeans to native talent finds no statistical support whatever in our study.
Finally, we must consider the impact of gender on interview rates and hiring (see Tables 11A-B). As the Committee also noted in its June 1999 and October 2000 report, females do tend to be interviewed at higher rates than males: in 2000-1, females averaged 2.57 interviews, males 2.10. As Table 11A shows, the gender difference in interview rates is even greater if one limits consideration to those candidates applying for more than ten or more than 20 positions. We cannot find any corollary factors that would explain the gender difference: women tend to apply for slightly fewer positions than men (62.6% of women at the annual meeting applied for more than ten positions, whereas 80.2% of men did), and women's strong presence in the field of art/archaeology (where few positions are available) would suggest that one might expect them to have fewer interviews than men. The fact that they nevertheless continue to do so well in obtaining interviews suggests that many departments are still actively attempting to address what they perceive as gender imbalances in their faculties. Table 11B shows that females also tend to be hired at a higher rate then males (57% vs. 43%), but that advantage appears to be entirely in non-permanent positions, since women and men are hired into tenured or tenure-track positions at comparable rates (19% vs. 20%).
Tables 11A-B also break down results by marital status: married individuals tend to be both interviewed and hired at somewhat higher rates than single individuals. Individuals who identify themselves as involved in non-marital long-term relationships appear to be at no real disadvantage in obtaining interviews, but are hired at a significantly lower rate. While it should not necessarily be assumed that those who place themselves in this category are individuals of minority sexual preference, some may be. On the other hand, the eleven candidates who chose to label themselves “minority sexual orientation” on the year-end questionnaire had an interview rate (2.36) close to the norm (2.28); seven of the eleven obtained new jobs, three of them tenure-track, somewhat better than the average rate of new hires and tenure-track hires for the general population of candidates. There is accordingly no conclusive evidence of discrimination based on sexual preference.
One common assumption that our results on marital status should help put to rest is the notion that married individuals are less willing to relocate than single candidates. Those who are geographically limited will tend to be among the group that applies for ten or fewer positions: as Table 11A shows, married individuals (or those involved in non-marital long-term relationships) are only slightly more prevalent among this group than among the population of job candidates as a whole: 57% of married candidates at the annual meeting applied for ten or fewer positions, whereas 51% of all candidates at the annual meeting did.
Tables 12A-B suggest that ethnic minorities do not experience any advantage in obtaining interviews or final hires. However, it should be cautioned that the numbers are too small to be definitive.
As we noted last year, publication does not necessarily correlate with success in obtaining interviews (see Table 13A). To this observation we can now add that it also has little positive correlation with obtaining jobs (see Table 13B), although having up to four articles published does correlate positively with the ability to obtain a permanent job. Male candidates appear to have published slightly more than female candidates: 26% of males and 20% of females at the annual meeting had a book published, 36% of males and 30% of females had three or more articles published. As we noted last year, publications tended to help male candidates more than females, in that males with books and up to nine articles obtained more interviews than males with no publications, whereas females with no publication tended to obtain more interviews than females with publication. To some extent this pattern continues in hiring, as indicated by Tables 13C-D. Clearly this phenomenon is one that needs further study over the coming years, but we find disturbing the conclusion that appears to emerge from our first two years of data: publication may be more expected of men than of women, and too much publication (i.e. a book or more than four articles) can hurt more than it helps.
Part III. Interview Experiences
Table 15 reflects candidate reaction to various issues which may have been raised inappropriately in interviews. The vast majority of candidates who completed this part of the year-end questionnaire reported no such issues being raised. The number who did is sufficiently small that it may have been a single institution in each case. The most common issue raised was religion, but some of the candidates noted in their written remarks that it was a religious institution, where it is of course entirely legal under both federal law and the Placement Guidelines to consider religion. Eight candidates said that they were asked about nationality, but again this is quite legal, as both U.S. and Canadian immigration law require that preference be given to qualified citizen candidates. Of more concern are the 15 candidates who were asked about marital status, the eleven who were asked about their partner’s willingness to relocate, and the nine who were asked about their children; these questions are in violation of the Placement Guidelines and we wish to remind all interviewers that they are inappropriate. We are happy that no candidates were asked about sexual orientation, physical health, or race and ethnicity. We do not find that any of these questions were particularly gender-discriminatory, except that males were somewhat more likely to be asked about religion, females about spousal relocation.
Many candidates added written comments to their questionnaires, for which the Committee is grateful. Any candidates with serious dissatisfaction over any aspect of the placement process, either related to institutional behavior or the operation of the Placement Service itself, is encouraged to contact the Committee on Placement directly. Most complaints can be handled with complete anonymity. In those few cases where it cannot be, the complainant will be so advised and asked for authorization before we proceed any further.
This report was written by Tom Hubbard who chaired the Placement Committee during 1997-2000. The Placement Committee wishes to thank Tom for his willingness to continue helping the committee in analyzing the data. We would also like to thank Barbara McManus, Vice President for Professional Matters, who has generated the necessary tables for the report and Irene Plonski, Director of the Placement Service, who collected the data.